Analysing the relationship between terrain constrained school accessibility and dropout patterns in West Kameng District Arunachal Pradesh India

Analysing the relationship between terrain constrained school accessibility and dropout patterns in West Kameng District Arunachal Pradesh India | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Analysing the relationship between terrain constrained school accessibility and dropout patterns in West Kameng District Arunachal Pradesh India Tushar Sarkar, Shobhit Pipil, Arghadeep Bose, Teresa Sakrinsow This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9210282/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 13 You are reading this latest preprint version Abstract Terrain is widely acknowledged as a barrier to schooling in mountainous regions. However, it is rarely modelled as an independent determinant of school accessibility and retention. This study examines the spatial relationship between terrain-induced accessibility and school-level dropout patterns across primary, middle, and secondary and above stages in West Kameng district, Arunachal Pradesh. A GIS-based school accessibility index (SAI) was constructed by integrating slope and land-use resistance within a cost-distance framework. The final analysis includes 41 primary, 58 middle, and 24 secondary and above schools. Accessibility remains valley concentrated. Most primary schools fall within moderate accessibility zones. None are located in very low-accessibility zones. However, dropout at the primary stage does not follow a consistent accessibility gradient. At the middle stage, the dropout rate increases as accessibility decreases. It increases from 3.24% in moderate zones to 9.31% in very low-accessibility zones. However, at the secondary level, the dropout rate remains the lowest in high-accessibility zones. Moderate and low-accessibility schools recorded relatively high rates of dropout. However, the pattern in this stage is less uniform. The results of the Kruskal-Wallis tests do not reveal statistically significant differences across SAI classes. However, descriptive gradients suggest that terrain-related accessibility constraints become more visible with educational progression. The findings indicate that terrain may function as a spatial conditioning factor rather than a direct determinant of dropout. Mobility support and residential planning are therefore critical for sustaining post-primary education in high-relief districts. Terrain-constrained accessibility School Accessibility Index (SAI) School dropout Spatial inequality Mountain regions Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1 Introduction Educational access in mountainous regions is shaped by multiple structural and spatial factors. Together, institutional availability, settlement dispersion, transport infrastructure, and household socioeconomic conditions influence school participation in such a setting. However, the physical landscape imposes a foundational constraint. The terrain limits movement, regulates connectivity, and conditions the effective reach of educational institutions [ 38 , 42 ]. Elevation, slope, surface roughness, and dense vegetation increase the effort required to traverse space [ 39 ]. For school-going children, this embodied difficulty often translates into irregular attendance. This often leads to early withdrawal from schooling [ 21 , 28 ]. Educational accessibility in mountainous contexts must therefore be understood as materially grounded in terrain-conditioned mobility. However, dropout outcomes are shaped by multiple social, economic, and institutional factors alongside physical accessibility. This cannot be explained by institutional proximity alone. Existing empirical research has largely examined the relationship between physical accessibility and schooling outcomes. Distance to school has been consistently identified as a critical determinant of enrolment and retention [ 11 , 21 , 24 ]. Evidence from Sub-Saharan Africa and South Asia shows that travel burdens affect schooling continuity across age groups [ 15 , 27 ]. In India, regional studies likewise link remoteness and infrastructural deficits to gaps in educational participation [ 17 ]. However, accessibility in most studies is measured through straight-line distance, travel time, or road connectivity. These measures capture proximity but not physical effort. Terrain is acknowledged but rarely operationalised as an independent explanatory factor. This limitation becomes more pronounced in mountainous environments where physical distance alone fails to represent movement difficulty. Studies in Himalayan regions have long documented the isolating effects of terrain on settlement systems and service delivery [ 22 ]. More recent work in Arunachal Pradesh has also highlighted terrain as a barrier to educational access [ 4 ]. However, quantitative modelling of terrain within formal accessibility frameworks remains limited. Advances in geographic information systems offer tools to address this gap [ 26 ]. Potential accessibility frames access as a function of spatial impedance rather than realised behaviour [ 19 ]. Cost-distance modelling incorporates terrain resistance through slope and land-cover surfaces [ 38 , 34 ]. Composite indices derived from such models have been widely used to evaluate access to health care and public services [ 14 ]. The applications of such indices to schooling systems remain relatively rare. This is more evident in high-relief regions of South Asia. Arunachal Pradesh provides a critical empirical context in this regard. The state is characterised by extreme topographic variability, dispersed settlements, and limited transport penetration. The West Kameng district exemplifies similar conditions. Settlements are often distributed across steep valleys and forested uplands where daily mobility is physically demanding. Schools may exist within administrative proximity, yet their effective accessibility varies sharply across terrain gradients. Against this backdrop, the present study examines terrain-conditioned accessibility within a GIS-based spatial framework. A school accessibility index (SAI) is constructed by integrating slope and land-use/land-cover resistance through cost-distance modelling. Accessibility conditions are evaluated across primary, middle, and secondary schooling stages. They are examined in relation to school-level dropout ratios. The analysis seeks to understand how terrain-induced accessibility varies spatially across the West Kameng district. It further explores whether these spatial accessibility patterns correspond with observed dropout patterns across school stages. 2 Theoretical Framework This study is situated within spatial accessibility theory. Accessibility is understood as a function of impedance rather than proximity [ 8 ]. The conceptual foundation derives from the potential accessibility formulation proposed by Hansen (1959) [ 19 ]. In this framework, accessibility reflects the resistance separating populations from opportunities. It represents structural possibility rather than realised behaviour. This perspective is particularly relevant in environments where travel data are unavailable. Accessibility can therefore be modelled as a property of spatial structure rather than individual mobility. However, the impedance is not uniform. It varies across physical landscapes. In mountainous regions, terrain constitutes a primary impedance field. Movement is structured by elevation gradients and slope morphology. Tobler (1993) [ 38 ] demonstrated that travel velocity changes systematically with gradient. Surface conditions introduce additional resistance. Vegetation density reduces permeability. Forest cover slows movement. Rivers and streams interrupt travel pathways. Ray & Ebener (2008) [ 34 ] reported that such barriers significantly alter accessibility patterns. Terrain impedance thus emerges as a composite function of topography and land cover. Cost-distance modelling operationalises this impedance structure within a spatial analytical framework [ 33 ]. It moves beyond Euclidean proximity. Instead, it estimates cumulative resistance across continuous terrain surfaces. Each spatial unit contributes incrementally to travel effort. Movement pathways adjust in response to resistance. Accessibility is therefore expressed as least accumulated effort rather than geometric distance. Such modelling has been widely applied in service accessibility research, including health and infrastructure planning [ 14 ]. Its application to schooling accessibility, particularly in mountainous regions, remains comparatively limited. Within this analytical structure, the school accessibility index functions as a composite accessibility metric. It integrates slope and land cover resistance within a unified cost surface. The index captures spatial variation in terrain-conditioned accessibility across school locations. These accessibility differentials reflect structural mobility gradients. Such gradients provide a spatial context for examining how terrain-related accessibility conditions correspond with schooling outcomes such as dropout. 3 Study Area West Kameng District lies in the western part of Arunachal Pradesh in northeastern India. It forms a part of the Eastern Himalayan mountain system. The district shares international boundaries with Bhutan and the Tibet Autonomous Region of China. Physiographically, it is characterised by rugged relief and steep slopes [ 5 ]. The valleys are deeply incised and spatially fragmented. The elevation varies sharply across short horizontal distances. This produces high terrain heterogeneity [ 25 ]. Dense forest cover further increases surface roughness. River gorges and unstable slopes increase movement difficulty. These geomorphic conditions impose strong constraints on mobility and spatial connectivity [ 37 ]. The settlement and service distribution in West Kameng reflect this terrain structure. The villages are small and spatially dispersed. Many are located along slopes and ridge corridors. Road penetration remains uneven across the district [ 9 ]. Foot-based mobility continues to play a significant role in everyday travel. Under such conditions, institutional proximity does not necessarily ensure functional access. Educational infrastructure operates within this terrain-constrained landscape. Students often traverse steep gradients and forested paths to reach schools [ 30 ]. Accessibility therefore varies sharply across short spatial distances. These characteristics make West Kameng an analytically appropriate setting for terrain-based accessibility modelling. 4 Database This study adopts a GIS-based potential accessibility framework to assess terrain-induced constraints on access to schools. Accessibility is treated as a spatial property of the landscape. It is not interpreted as observing student travel behaviour. The analysis focuses on the relative physical difficulty imposed by existing terrain conditions. Terrain has been identified as an important constraint to educational access in Arunachal Pradesh [ 4 , 2 ]. In this study, terrain is operationalised explicitly within the accessibility framework as a conditioning factor. Other determinants of dropout such as household socio-economic conditions, institutional quality, and cultural factors are acknowledged but are not directly modelled in this spatial analysis. The framework follows the concept of potential accessibility proposed by Hansen (1959) [ 19 ]. Here, accessibility is defined by spatial impedance. This approach is widely used in geography and spatial planning [ 36 ]. This approach is suitable when reliable data on student residences and daily travel routes are unavailable [ 19 , 31 ]. Terrain-induced accessibility is represented using slope and land-use/land-cover variables. Slope captures the dominant topographic constraint on walking. However, land cover represents surface-related barriers such as vegetation density and surface roughness. Together, these variables capture the primary physical properties of terrain relevant to pedestrian movement in mountainous environments. The school accessibility index (SAI) follows established practices in GIS-based accessibility modelling [ 3 , 29 ]. Cost-distance approaches have been used to represent spatial impedance through resistance surfaces [ 38 , 34 ]. Normalised composite indices have been applied to compare access to services such as schools and health facilities [ 14 ]. The SAI builds on this tradition. It provides a relative and comparable measure of terrain-induced accessibility across schools and educational stages. 4.1 School location data School locations constitute the primary spatial input for the study. A total of 209 schools located in West Kameng district were identified from the Unified District Information System for Education Plus (UDISE+) Schools Geo-Portal. A ready-to-use GIS point layer is not available from any official sources. Therefore, school locations have been manually digitised using Google Earth Pro. Each school has been identified using high-resolution satellite imagery. Locations have been cross-verified with information available on the UDISE+ portal. Each digitised school has been assigned a unique school ID corresponding to the official UDISE code. Schools have been classified according to the highest-grade category offered for analytical standardisation. All primary schools have been treated as offering education up to Class V. All middle schools have been treated as offering education up to Class VIII. Schools offering education from Class IX onwards have been classified as secondary and above. This standard has been adopted solely to define school categories and associated buffer distances. This does not imply uniformity in the actual grade structure across schools. 4.2 School attribute data School-level attribute data were compiled from Department of School Education, Government of Arunachal Pradesh. These data include total enrolment and total dropout. The data represent combined enrolment and dropout figures for the 2023-24 and 2024-25 academic years. The two years were combined to obtain a stable school-level dropout ratio. Combining consecutive academic years reduces year-specific fluctuations. Dropout has been operationalised as a school-level dropout ratio rather than an absolute count. The dropout ratio for each school was calculated following the procedure outlined by the Government of India (2023) [ 16 ] as follows: Dropout ratio = \(\:\:\frac{Total\:dropout\:in\:a\:year}{Total\:enrolment\:at\:the\:begining\:of\:the\:year}\times\:100\) This formulation standardises dropout across schools with different enrolment sizes. It enables comparisons across accessibility classes. Individual student trajectories have not been analysed. The attribute data were linked to the spatial school layer using the School ID. Schools lacking enrolment records were excluded from the analysis. Most missing data reflect administrative restructuring in the district. Several schools were merged with nearby institutions. Some were closed owing to persistently low enrolment. These changes are not systematically related to accessibility conditions. Among the 209 identified schools, 86 institutions (41.15%) were excluded for these reasons. The final analytical sample therefore includes 41 primary schools, 58 middle schools, and 24 secondary and above schools. This retained sample remains adequate for examining relative terrain-induced accessibility effects rather than absolute dropout levels. 4.3 Digital elevation model (DEM) Topographic data were obtained from the Copernicus Sentinel Digital Elevation Model (DEM). The DEM has a spatial resolution of 10 metres. It has been selected to represent rugged mountainous terrain. The resolution has been found to be compatible with other datasets used in the study. The DEM has been projected to WGS 84, UTM Zone 46 N, and has been clipped to the administrative boundary of West Kameng district. 4.4 Land use/land cover data The land use and land cover data were obtained from the ESA WorldCover Version 2 (2021) dataset with a 10-meter spatial resolution [ 40 ]. The dataset provides a consistent classification of surface cover types. 5 Methodology 5.1 Spatial harmonisation and preprocessing All spatial datasets were projected to a common coordinate reference system, WGS 84, UTM Zone 46 N. The raster layers were aligned to a spatial resolution of 10 metres. Consistent snap raster and extent settings have been applied. This approach ensures pixel-level correspondence across raster layers. 5.2 Slope derivation and terrain resistance modelling The slope is derived from the DEM and represented in degrees. The slope angle provides a direct indicator of walking difficulty in mountainous terrain [ 1 ]. The continuous slope surface was reclassified into five categories representing increasing levels of movement difficulty. The slope classification follows thresholds commonly used in land evaluation and geomorphological studies, as formalised in the FAO (1976) [ 12 ] land evaluation guidelines. These thresholds group slopes according to increasing physical limitations to human activity. The classes have been adapted to reflect pedestrian movement difficulty in the Himalayan terrain. Each slope class was assigned an ordinal resistance value ranging from 1 to 5. The resistance values represent the relative walking difficulty. They do not represent the calibrated physical cost. Table 1 Slope classes and resistance values used in terrain accessibility modelling Slope class Slope range (degrees) Terrain interpretation Resistance value Very gentle 20° Severe difficulty 5 5.3 Land cover resistance modelling The land cover resistance was modelled using the WorldCover dataset [ 41 ]. Each land-cover class was evaluated for pedestrian movement difficulty. The classes have been reclassified into resistance values ranging from 1–5 [ 29 ]. The same ordinal scale used for slope resistance was applied to maintain conceptual consistency. Dense forest and water bodies were assigned the maximum resistance values. These classes represent severe constraints on movement. This approach avoids introducing absolute barriers while retaining realistic terrain impedance. Table 2 Land-use/land-cover classes and resistance values used in accessibility modelling Land-cover class Description Resistance value Accessibility implication Built-up Settlements, roads 1 Very easy movement Cropland / Grassland Agricultural land 2 Easy movement Shrubland Secondary vegetation 3 Moderate obstruction Tree cover Open or mixed forest 4 Difficult movement Moss and lichen Alpine forest, sparse ground cover 4 Difficult movement due to cold, moisture, uneven surface Dense forest Closed canopy forest 5 Very difficult movement Water bodies Rivers, lakes 5 Severe constraint Snow and ice Permanent snow, glaciers 5 Severe constraint; often impassable 5.4 Construction of the composite cost surface Slope resistance and land cover resistance have been integrated to create a composite cost surface [ 20 , 18 ]. Both components have been assigned equal weights. The composite cost was calculated as the linear aggregation of the two resistance values. Equal weighting has been adopted deliberately. Resistance values are ordinal indicators of relative movement difficulty. They do not represent the calibrated physical cost. Empirical data quantifying the differential contributions of slope and land cover are not available. Equal weighting therefore provides a conservative modelling choice. This approach avoids introducing subjective bias. The composite surface represents the cumulative terrain-induced constraint rather than the absolute travel cost. Sensitivity testing of alternative weighting schemes has not been conducted because of the absence of calibrated pedestrian movement data for the district. 5.5 Cost-distance modelling Cost-distance modelling has been employed to estimate potential physical accessibility under terrain-induced constraints [ 23 ]. Unlike Euclidean distance, cost-distance analysis accounts for spatial heterogeneity in movement difficulty. It incorporates resistance values derived from terrain and surface conditions. In mountainous regions, the straight-line distance poorly represents the actual movement effort. Therefore, cost-distance modelling provides a more realistic approximation of terrain-conditioned accessibility [ 10 ]. Using the composite cost surface, cost‒distance analysis was performed with schools treated as destination points. The analysis calculates the least cumulative cost required to reach the nearest school from each raster cell. The resulting raster represents potential physical accessibility rather than observed travel behaviour. 5.6 Delineation of school-centred accessibility buffers The student residence data are not available from official sources. School-centred buffers have therefore been used to approximate potential catchment areas. Buffer distance is defined according to the educational stage and the expected daily mobility of students. A 1 km, 2 km, and 3 km buffer has been applied to primary, middle, and secondary and above schools, respectively [ 13 ]. These buffers represent potential catchment extents. They approximate potential accessibility zones rather than actual travel routes. 5.7 Extraction of school-level accessibility measures Zonal statistics were computed in ArcGIS Pro using Zonal Statistics as a table tool. School buffer polygons were used as zones. The cost distance raster served as the input value raster. For each buffer, the mean cost distance value is extracted. This value represents the average terrain-induced movement effort within the potential catchment of each school. The extracted values have been linked back to the school attribute table using the School ID. 5.8 Construction of the School Accessibility Index (SAI) The school accessibility index (SAI) was calculated using min-max normalisation of the mean cost-distance values. This follows established GIS-based accessibility indices in educational geography [ 32 ]. It has been performed separately for primary, middle, and secondary and above schools. This accounts for differences in buffer size and accessibility context across educational stages. The school accessibility index (SAI) is calculated as follows: SAI i = \(\:\frac{{C}_{i}-{C}_{min}}{{C}_{max}-{C}_{min}}\) Here, C i represents the mean cost-distance value for school i. Cₘ i ₙ and Cₘₐₓ represent the minimum and maximum values within the same school category. Lower SAI values indicate greater accessibility. Higher values indicate greater terrain-induced constraints. The SAI values are grouped into four accessibility zones. The classification was carried out separately for each school category. Table 3 School Accessibility Index (SAI) classification scheme SAI range Accessibility class Interpretation 0.00–0.25 High accessibility Minimal terrain-induced constraint 0.26–0.50 Moderate accessibility Low to moderate terrain constraint 0.51–0.75 Low accessibility High terrain-induced constraint 0.76–1.00 Very low accessibility Severe terrain-induced constraint 5.9 Statistical analysis of accessibility and dropout The relationship between terrain-induced accessibility and school dropout has been examined using nonparametric statistical methods. These methods are appropriate because of their small sample sizes, nonnormal distributions, and the presence of tied values [ 35 , 6 ]. Spearman’s rank correlation was initially considered to assess the monotonic association between the school accessibility index (SAI) and the school-level dropout ratio. The SAI is treated as an ordinal variable. However, the correlation could not be reliably estimated because of extensive tied ranks in both accessibility classes and dropout ratios. Therefore, Kruskal-Wallis H tests were applied to examine differences in dropout ratios across SAI classes. The tests were conducted separately for primary, middle, and secondary schools and above. 6 Results and Discussion 6.1 Terrain resistance structure of West Kameng district Slope and land-use resistance together produce a highly constrained mobility surface across the West Kameng district. Steep gradients dominate the landscape (Fig. 3 ; Table 4 ). The highest slope resistance class (very steep) alone accounts for 76.12% of the total geographical area of the district (Table 4 ). In contrast, very gentle and gentle slopes together form only 2.41% of the district’s total area. This indicates that most settlements are embedded within physically demanding terrain. Daily movement therefore requires greater time and effort. This suggests that school travel becomes physically burdensome in such environments. Ives & Messerli (2003) [ 22 ] documented a similar terrain‒mobility constraint across Himalayan mountain systems. They noted steep relief fragment settlement connectivity and limited service accessibility in mountainous regions. Table 4 Areal distribution of slope resistance classes Slope resistance value Slope class Area (sq. km) % Area 1 Very gentle 62.60 0.84 2 Gentle 116.99 1.57 3 Moderate 252.75 3.41 4 Steep 1339.69 18.05 5 Very steep 5649.97 76.12 Land cover resistance deepens this constraint. A large share of the district is covered by high-resistance surfaces. These mainly include tree cover (open and mixed forest), alpine forest, and areas with sparse ground cover. Together, they account for 91.56% of the district’s total area (Fig. 4 ; Table 5 ). In contrast, low-resistance surfaces such as built-up (0.16%) and cropland (8%) land remain spatially fragmented. These pockets rarely form continuous access corridors. The composite terrain classification reflects this cumulative resistance. Nearly 87.27% of the district falls within the severe terrain difficulty zone (Table 6). In contrast, very easy and easy terrains together account for less than 2%. Accessibility is therefore structurally valley bound and ridge constrained. Institutional reach becomes uneven. The terrain therefore forms the physical context within which spatial differences in school accessibility emerge. Similar cumulative resistance effects have been reported in accessibility modelling across South Asian and Southeast Asian Mountain regions [ 39 ]. In such terrains, slope and surface impedance operate together to constrain movement. Table 5 Areal distribution of land-use/land-cover resistance classes LULC resistance Area (sq. km) % Area 1 11.72 0.16 2 593.70 8.00 3 0.0086 0.0001 4 6795.23 91.56 5 21.33 0.29 Table 6 A real distribution of composite terrain resistance zones Zone Area (sq. km) % Area Very Easy 21.21 0.29 Easy 111.27 1.50 Moderate 248.72 3.35 Difficult 563.76 7.59 Severe 6477.03 87.27 6.2 Composite terrain cost surface of West Kameng district The composite terrain cost surface integrates slope steepness and land cover resistance into a continuous representation of cumulative movement difficulty. High-cost terrain dominates almost the entire district (Fig. 5 ). This pattern is particularly evident across the upland and dissected interior regions. These areas coincide with high mountain ridges and densely forested slopes. The terrain is structurally dissected. Movement across such landscapes requires considerable physical effort. Travel time may also increase under such terrain conditions. Seasonal conditions such as landslides or surface wetness could further intensify these constraints. This ridge-concentrated resistance pattern aligns with the anisotropic cost-distance modelling approaches applied in mountainous regions. Studies from high-relief landscapes, including parts of South Asia and Southeast Asia, have shown that slope acts as a directional impedance. It alters effective movement pathways and increases cumulative travel effort. Similar modelling applications have demonstrated how ridge systems restructure accessibility surfaces in terrain-dominated environments [ 38 , 34 ]. Low-cost corridors occur only in fragmented pockets. These belts are largely confined to valley floors and river-aligned settlement belts (Fig. 5 ). The Kameng River valley and its tributary systems form the corridors with the most visible accessibility. The terrain in these zones appears relatively navigable. Settlement clustering and road presence are also more evident. Schools located within such corridors are likely to have relatively wider functional catchments. In contrast, institutions situated along ridge slopes or forested uplands remain embedded within high cumulative cost zones. Their accessibility reach is likely to remain spatially restricted. The composite surface therefore suggests that terrain-induced mobility gradients structure spatial differences in potential school accessibility across the district. 6.3 Spatial inequality in accessibility Accessibility inequality across West Kameng District follows a clear elevation gradient. Primary schools remain heavily concentrated below 2000 metres (Fig. 6 ; Table 7 ). The 1000–2000 metre belt alone accounts for 29 (70.73%) schools out of the 41 total primary schools. This finding indicates that the lower-altitude valley system forms the main institutional zone at the foundational stage. Within this belt, 51.72 percent fall within moderate-accessibility zones, and 27.59 percent fall within high-accessibility zones (Table 7 ). However, another 20.69 percent of the cases occur in low-accessibility terrain. However, it is worth noting that no primary school occurs within the very low accessibility category at any elevation. This pattern indicates that foundational schooling has largely been positioned within terrain environments that remain physically manageable for younger children. Accessibility inequality at the middle stage broadly follows the primary pattern. However, it begins to expand into more difficult terrain. The 1000–2000 meter belt also recorded a high concentration of middle schools, which closely mirrors the distribution of primary schools within this elevation range. This belt accounted for 39 schools (67.24%) out of the total 58 Table 7 Elevation-wise distribution of primary schools across accessibility (SAI) classes Elevation (Meter) SAI class Total High accessibility Moderate accessibility Low accessibility Very low accessibility < 1000 1 (20.00) 2 (40.00) 2 (40.00) 0 (0.00) 5 (100.00) 1000–2000 8 (27.59) 15 (51.72) 6 (20.69) 0 (0.00) 29 (100.00) 2001–3000 1 (14.29) 4 (57.14) 2 (28.57) 0 (0.00) 7 (100.00) 3001–4000 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) > 4000 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) Total 10 (24.29) 21 (51.22) 10 (24.39) 0 (0.00) 41 (100) Note : Figures in parentheses are percentages Table 8 Elevation- distribution of middle schools across accessibility (SAI) classes Elevation (Meter) SAI class Total High accessibility Moderate accessibility Low accessibility Very low accessibility < 1000 3 (42.86) 3 (42.86) 1 (14.29) 0 (0.00) 7 (100.00) 1000–2000 14 (35.90) 19 (48.72) 4 (10.26) 2 (5.13) 39 (100.00) 2001–3000 3 (25.00) 5 (41.67) 3 (25.00) 1 (8.33) 12 (100.00) 3001–4000 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) > 4000 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) Total 20 (34.48) 27 (46.55) 8 (13.79) 3 (5.17) 58 (100.00) Note : Figures in parentheses are percentages middle schools in the district (Fig. 7 ; Table 8 ). This confirms the continued concentration within lower-altitude settlement corridors. Within this belt, 48.72% of the schools fall in moderate-accessibility zones, and 35.90% fall in high-accessibility zones (Table 8 ). However, 15.39% of cases occur in low- and very low-accessibility terrains. Terrain exposure therefore widens beyond the foundational stage. The shift becomes clearer above 2000 metres. In the 2001–3000 m belt, 25.00% of the transects fall in high-accessibility zones. However, 33.33% of the schools operate under low- and very-low-accessibility conditions. Compared with primary schools, middle institutions show greater terrain exposure. This gradient becomes more pronounced at the secondary level and above. In contrast to the primary and middle stages, the concentration of schools declines in this tier. The 1000–2000 m belt contains 14 out of the 24 secondary schools and above, accounting for 58.33% of the total (Fig. 8 ; Table 9 ). This indicates partial expansion beyond the core valley corridors. Within this belt, 64.29% of the area remains in high-accessibility terrain (Table 9 ). Conditions become more challenging beyond 2000 metres. In the 2001–3000 m belt, only 12.50% of the transects fall in high-accessibility zones. However, 50.00% of the cases occur in low- and very low-accessibility environments. A clear progression therefore emerges. Primary schools remain valley-buffered. The middle schools have moderate terrain exposure. Secondary institutions and higher-level institutions face the greatest constraints. Accessibility inequality therefore becomes more pronounced with educational progression and elevation. Table 9 Elevation wise distribution of secondary and above school across accessibility (SAI) classes Elevation (Meter) SAI class Total High accessibility Moderate accessibility Low accessibility Very low accessibility < 1000 1 (50.00) 0 (0.00) 0 (0.00) 1 (50.00) 2 (100.00) 1000–2000 9 (64.29) 4 (28.57) 1 (7.14) 0 (0.00) 14 (100.00) 2001–3000 1 (12.50) 3 (37.50) 1 (12.50) 3 (37.50) 8 (100.00) 3001–4000 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) > 4000 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) 0 (0.00) Total 11 (45.83) 7 (29.17) 2 (8.33) 4 (16.67) 24 (100.00) Note : Figures in parentheses are percentages 6.4 Dropout patterns across accessibility zones in West Kameng district Dropout patterns vary across accessibility classes. However, the strength of the association differs by schooling stage. At the primary level, the pattern is irregular. The mean dropout rate is highest in high-accessibility zones, at 11.09% (Table 10 ). It decreases sharply in moderate accessibility areas to 3.72%. It rises again in low-accessibility terrain to 9.65%. No school falls within the very low accessibility class. This absence of a consistent gradient suggests that terrain constraint does not appear to strongly structure early-stage withdrawal from schooling. At this stage, household conditions or schooling readiness may play a greater role than travel burden alone. Huisman & Smits (2009) [ 21 ] reported a similar pattern in mountainous regions. Their findings suggest that household vulnerability often has a stronger effect than travel distance in shaping early-grade retention and dropout. Table 10 Distribution of primary schools and mean dropout ratio across accessibility (SAI) classes SAI class No. of schools Percentage of school Mean dropout ratio (%) High accessibility 10 24.39 11.09 Moderate accessibility 21 51.22 3.72 Low accessibility 10 24.39 9.65 Very low accessibility 0 0.00 - All primary schools 41 100.00 6.96 The influence of the terrain constraint becomes more visible beyond the primary level. At the middle stage, the dropout rate increases steadily as accessibility decreases. It remains lowest in moderate-accessibility zones at 3.24% (Table 11 ). It increases in high-accessibility areas to 5.42%. It increases further in low-accessibility terrain to 7.73%. The highest dropout rate of 9.31% occurs in very low-accessibility environments (Table 11 ). This progression suggests that terrain-related travel effort may become more relevant during the transition from primary to middle schooling. Such distance-related discontinuation at post primary stages has also been documented across rural schooling systems in developing regions [ 27 ]. The secondary and above stage shows a moderated but related pattern. Dropout remains lowest in high-accessibility zones at 3.44% (Table 12 ). It increases in moderate- and low-accessibility environments to 9.17% and 8.44%, respectively. Very low-accessibility schools record a dropout ratio of 5.35 percent (Table 12 ). Despite this variation, higher dropout levels remain concentrated outside the high-accessibility class. This partial gradient reflects findings from South Asian hill regions, where continuation beyond elementary schooling becomes mobility sensitive rather than proximity dependent [ 17 ]. Taken together, terrain accessibility does not produce a uniform dropout response. Its role appears weak at the primary stage. A clearer descriptive gradient emerges during middle schooling and remains partially visible at the secondary and above level. As students progress to higher grades, commuting routes often become longer. Daily travel may also become more physically demanding. In such settings, terrain-related travel burden may interact with household and institutional conditions that influence school continuation. Terrain therefore does not directly determine dropout. Instead, it forms part of the broader spatial context within which withdrawal from schooling may occur. Table 11 Distribution of middle schools and mean dropout ratio across accessibility (SAI) classes SAI class No. of schools Percentage of school Mean dropout ratio (%) High accessibility 20 34.48 5.42 Moderate accessibility 27 46.55 3.24 Low accessibility 8 13.79 7.73 Very low accessibility 3 5.17 9.31 All middle schools 58 100.00 4.93 Table 12 Distribution of secondary and above schools and mean dropout ratio across accessibility (SAI) classes SAI class No. of schools Percentage of school Mean dropout ratio (%) High accessibility 11 45.83 3.44 Moderate accessibility 7 29.17 9.17 Low accessibility 2 8.33 8.44 Very low accessibility 4 16.67 5.35 All secondary and above schools 24 100.00 5.84 6.5 Statistical relationship between accessibility and dropout Nonparametric tests have been applied to examine whether dropout ratios vary significantly across accessibility classes. The Kruskal‒Wallis results show limited statistical differentiation. The H values of primary, middle and secondary and above schools were 0.457 (p = 0.796), 2.611 (p = 0.456) and 0.698 (p = 0.874), respectively (Table 13 ). In all the cases, the p values remain above the 0.05 threshold. Therefore, dropout variation across SAI classes is not statistically significant within the present dataset. However, descriptive patterns reveal a more nuanced gradient pattern. Dropout ratios generally rise as accessibility declines, particularly beyond the primary stage. Middle schools show greater dropout rates within low- and very-low-accessibility zones. Secondary and above schools also recorded comparatively higher dropout levels outside high-accessibility classes. These descriptive patterns suggest that terrain-related accessibility may operate as a contextual constraint rather than a statistically isolated predictor. The absence of statistical significance may reflect structural data limitations. Extreme accessibility categories contain fewer schools. This reduces the statistical power of nonparametric tests. The analysis is also conducted at the school level rather than at the level of individual student residences. Consequently, the terrain conditions experienced by individual students may not be fully captured by school-centred accessibility measures. Terrain accessibility may therefore operate as a conditioning factor rather than a direct causal determinant. Its influence is likely to interact with institutional reach, travel fatigue, and stage-specific mobility demands. Accessibility inequality nevertheless remains analytically meaningful. In such terrain-dominated environments, spatial accessibility differences provide an important contextual dimension for understanding variation in schooling continuation. Table 13 Kruskal–Wallis test results for differences in dropout ratios across accessibility (SAI) classes by school stage Stage of School H Statistics p-value Primary 0.457 0.796 Middle 2.611 0.456 Secondary & above 0.698 0.874 7 Conclusions This study examined terrain-constrained physical accessibility to schools in West Kameng district using a GIS-based school accessibility index. It also explored how accessibility conditions are related to school-level pattern of dropout across educational stages. The findings indicate that terrain forms a structurally embedded constraint on educational access. Steep slopes dominate the district surface. Dense forest cover further intensifies movement difficulty. Severe terrain difficulty characterises most geographical areas. Accessibility therefore remains concentrated within valley corridors. Ridge and upland locations experience restricted mobility conditions. The school distribution reflects this terrain structure. Primary institutions remain largely valley-buffered. Most are located within moderate-accessibility environments. Terrain exposure increases during the middle and secondary stages. Higher-level schools extend into more physically demanding landscapes. Dropout patterns show a differentiated relationship. The influence of the terrain appears weak at the primary stage. A clearer descriptive gradient emerges during middle schooling. The pattern continues at the secondary and above level, although less uniformly. As students progress to a higher level of education, commuting routes often become longer. Daily travel may therefore become increasingly demanding under steep terrain conditions. The accessibility burden appears to accumulate with educational progression. The statistical tests do not reveal significant variation in dropout ratios across accessibility classes. However, descriptive patterns suggest that higher dropout levels are more frequently observed outside high-accessibility zones. This is more evident particularly beyond the primary stage of schooling. These findings indicate that terrain accessibility does not operate as an isolated determinant of dropout. Instead, it forms part of the broader spatial context within which schooling continuation takes place. In mountainous environments, mobility constraints may interact with institutional reach, household vulnerability, and travel fatigue. Terrain therefore acts as a conditioning environment that may amplify existing educational disadvantages rather than directly producing dropout. This study contributes by modelling terrain-conditioned accessibility within a GIS-based framework. The school accessibility index provides a stage-sensitive measure of accessibility differences across schools in mountainous terrain. The results suggest that accessibility inequalities become more visible beyond foundational schooling. These findings carry policy implications. Educational expansion in high-relief regions must consider terrain-conditioned mobility rather than administrative proximity alone. Residential schooling and transport support may therefore be important for sustaining post-primary education. Future research may incorporate household location data and actual student travel routes. Integrating socio-economic and institutional variables would further clarify how terrain accessibility interacts with broader determinants of school participation. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Funding This research did not receive any funding from any individual or organization. Author Contribution First and corresponding author conceptualized the study and prepared the original draft of the manuscript. First co-author prepared the maps using GIS and provided methodological support. Second co-author contributed to the methodological framework and supported discussion. 3rd co-author assisted in data collection and compilation. Data Availability The datasets used in this study are publicly available. School-level data were obtained from the [UDISE+](https:/schoolgis.nic.in) portal. Spatial datasets include the [Copernicus Digital Elevation Model (DEM)](https:/dataspace.copernicus.eu/explore-data/data-collections/copernicus-contributing-missions/collections-description/COP-DEM) and [ESA WorldCover](https:/esa-worldcover.org/en/data-access) (2021). Processed data may be made available from the corresponding author upon reasonable request. References Abdul Yamin NAA, Basaruddin KS, Abu Bakar S, Salleh AF, Som M, Yazid MH, H., Hoang TD. (2022). Quantification of gait stability during incline and decline walking: The responses of required coefficient of friction and dynamic postural index. Journal of Healthcare Engineering, 2022(1), 7716821. https://doi.org/10.1155/2022/7716821 Alam S. (2024, January 10). The obstacles to Arunachal's education system. The Arunachal Times. https://arunachaltimes.in/index.php/2024/01/10/the-obstacles-to-arunachals-education-system/ (Accessed 20 February 2026). Al-Sabbagh TA. GIS location-allocation models in improving accessibility to primary schools in Mansura. city-Egypt Geoj. 2022;87(2):1009–26. https://doi.org/10.1007/s10708-020-10290-5 . Biswas G, Sharma MG. Impact of geographical factors on education in Papum Pare District of Arunachal Pradesh. Res J Educational Sci. 2018;6(2):1–5. https://oldror.lbp.world/UploadedData/6740.pdf . Central Ground Water Board. (2013). Ground water information booklet: West Kameng District, Arunachal Pradesh. Ministry of Water Resources, Government of India. https://www.cgwb.gov.in/old_website/District_Profile/Arunachal/West%20Kameng.pdf (Accessed 10 February 2026). Conover WJ. Practical nonparametric statistics. 3rd ed. Wiley; 1999. Copernicus Programme. (2021). Copernicus digital elevation model (DEM) 10 m. European Space Agency. https://doi.org/10.5270/ESA-c5d3d65 Damurski Ł. Neighbourhood cohesion and territorial cohesion: in search for conceptual integrity. GeoJournal. 2022;87(6):4635–51. https://doi.org/10.1007/s10708-021-10523-1 . Development, Commissioner, Government of India. Ministry of Micro, Small & Medium Enterprises,. (2014). Brief industrial profile of West Kameng District. Government of India. https://dcmsme.gov.in/dips/DIP_West-Kameng-Dist.pdf (Accessed 7 February 2026). Druga M, Minár J. Cost distance and potential accessibility as alternative spatial approximators of human influence in LUCC modelling. Land Use Policy. 2023;132:106840. https://doi.org/10.1016/j.landusepol.2023.106840 . Filmer D. If you build it, will they come? School availability and school enrolment in 21 poor countries. J Dev Stud. 2007;43(5):901–28. https://doi.org/10.1080/00220380701384598 . Food and Agriculture Organization of the United Nations. (1976). A framework for land evaluation (FAO Soils Bulletin No. 32). FAO. https://www.fao.org/4/x5310e/x5310e00.htm (Accessed 12 October 2025). Gagnon A. (2022). Isochrone-based catchment areas for educational planning. International Institute for Educational Planning (IIEP), UNESCO. https://www.iiep.unesco.org/en/publication/isochrone-based-catchment-areas-educational-planning (Accessed 1 January 2026). Geurs KT, van Wee B. Accessibility evaluation of land-use and transport strategies: Review and research directions. J Transp Geogr. 2004;12(2):127–40. https://doi.org/10.1016/j.jtrangeo.2003.10.005 . Glewwe P, Jacoby HG. An economic analysis of delayed primary school enrollment in a low-income country: The role of early childhood nutrition. Rev Econ Stat. 1995;77(1):156–69. https://www.jstor.org/stable/2110001 . Government of India, Ministry of Education, Department of School Education and Literacy. (2023). School dropout rate: Concept, formula and procedure. Government of India. https://educationforallinindia.com/wp-content/uploads/2023/01/Drop-out-rate-formula.pdf (Accessed 13 October 2026). Govinda R, Bandyopadhyay M. (2008). Access to elementary education in India: Country analytical review. https://eric.ed.gov/?id=ED508807 Grogan A. (2009). Creating a spatial analysis model for generating composite cost surfaces to depict cross country mobility in natural terrain. In ASPRS/MAPPS Fall Conference. Texas. Hansen WG. How accessibility shapes land use. J Am Inst Planners. 1959;25(2):73–6. https://doi.org/10.1080/01944365908978307 . Hengl T, Reuter HI. Geomorphometry: A brief guide. In: Hengl T, Reuter HI, editors. Geomorphometry: Concepts, software, applications. Elsevier; 2009. pp. 1–33. Huisman J, Smits J. Effects of household- and district-level factors on primary school enrolment in 30 developing countries. World Dev. 2009;37(1):179–93. https://doi.org/10.1016/j.worlddev.2008.01.007 . Ives JD, Messerli B. The Himalayan dilemma: Reconciling development and conservation. Routledge; 2003. Jobe RT, White PS. A new cost-distance model for human accessibility and an evaluation of accessibility bias in permanent vegetation plots in Great Smoky Mountains National Park, USA. J Veg Sci. 2009;20(6):1099–109. Kellerman A. Distance: A 2024 re-examination. GeoJournal. 2024;89(4):183. https://doi.org/10.1007/s10708-024-11176-6 . Krishi Vigyan Kendra West Kameng. (2024). District profile. District profile. Krishi Vigyan Kendra, West Kameng https://kvkwestkameng.gov.in/districtprofile.html (Accessed 12 January 2026). Liu L, Li X. GIS and built environment. GeoJournal. 2014;79(2):127–8. https://doi.org/10.1007/s10708-013-9507-9 . Lloyd CB, Mensch BS, Clark WH. The effects of primary school quality on school dropout among Kenyan girls and boys. Comp Educ Rev. 2000;44(2):113–47. https://doi.org/10.1086/447600 . Macharia PM, Mumo E, Okiro EA. Modelling geographical accessibility to urban centres in Kenya in 2019. PLoS ONE. 2021;16(5):e0251624. https://doi.org/10.1371/journal.pone.0251624 . Meena DK, Tripathi R, Agrawal S. An evaluation of primary schools and their accessibility using GIS techniques: A case study of Prayagraj district. India Geoj. 2023;88(2):1921–51. https://doi.org/10.1007/s10708-022-10715-3 . Ministry of Education, Government of India. (2022). Unified District Information System for Education Plus (UDISE+) 2021–22: All India report card on school education. Government of India. https://www.education.gov.in/sites/upload_files/mhrd/files/statistics-new/udise_21_22.pdf (Accessed 11 September 2025). Neutens T. Accessibility, equity and health care: Review and research directions for transport geographers. J Transp Geogr. 2015;43:14–27. https://doi.org/10.1016/j.jtrangeo.2014.12.006 . O'Sullivan D, Morrison A, Shearer J. Using desktop GIS for the investigation of accessibility by public transport: An isochrone approach. Int J Geogr Inf Sci. 2000;14(1):85–104. https://doi.org/10.1080/136588100240976 . Pallathadka A, Pallathadka L, Rao S, Chang H, Van Dommelen D. Using GIS-based spatial analysis to determine urban greenspace accessibility for different racial groups in the backdrop of COVID-19: A case study of four US cities. GeoJournal. 2022;87(6):4879–99. https://doi.org/10.1007/s10708-021-10538-8 . Ray N, Ebener S. AccessMod 3.0: Computing geographic coverage and accessibility to health care services using anisotropic movement of patients. Int J Health Geogr. 2008;7(1):63. https://doi.org/10.1186/1476-072X-7-63 . Siegel S. Nonparametric statistics for the behavioral sciences. McGraw-Hill; 1956. Soukhov A, Pereira RHM, Higgins CD, Páez A. A family of accessibility measures derived from spatial interaction principles. PLoS ONE. 2025;20(11):e0335951. https://doi.org/10.1371/journal.pone.0335951 . Tiwari PC, Tiwari A, Joshi B. Urban growth in Himalaya: Understanding the process and options for sustainable development. J Urban Reg Stud Contemp India. 2018;4(2):15–26. https://doi.org/10.15027/45582 . Tobler W. (1993). Three presentations on geographical analysis and modeling: Non-isotropic geographic modeling; speculations on the geometry of geography; and global spatial analysis (NCGIA Technical Report 93 – 1). National Center for Geographic Information and Analysis, University of California, Santa Barbara. https://escholarship.org/uc/item/05r820mz Weiss, D. J., Nelson, A., Gibson, H. S., Temperley, W., Peedell, S., Lieber, A., …Gething, P. W. (2018). A global map of travel time to cities to assess inequalities in accessibility. Nature, 553(7688), 333–336. https://doi.org/10.1038/nature25181. Zanaga D, Van De Kerchove R, Daems D, De Keersmaecker W, Brockmann C, Kirches G, Wevers J, Cartus O, Santoro M, Fritz S, Lesiv M, Herold M, Tsendbazar NE, Xu P, Ramoino F, Arino O. (2022). ESA WorldCover 10 m 2021 v200 [Data set]. Zenodo. https://doi.org/10.5281/zenodo.7254221 . Zanaga, D., Van De Kerchove, R., De Keersmaecker, W., Souverijns, N., Brockmann, C.,Quast, R., … Arino, O. (2021). ESA WorldCover 10 m 2020 v100 (pp. 1–27). https://doi.org/10.5281/zenodo.5571936. Zhu Y, Zinda JA, Liu Q, Wang Y, Fu B, Li M. Accessibility of primary schools in rural areas and the impact of topography: A case study in Nanjiang County, China. Land. 2023;12(6):1134. https://doi.org/10.3390/land12061134 . Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 17 Apr, 2026 Reviews received at journal 08 Apr, 2026 Reviews received at journal 06 Apr, 2026 Reviews received at journal 05 Apr, 2026 Reviewers agreed at journal 04 Apr, 2026 Reviewers agreed at journal 01 Apr, 2026 Reviewers agreed at journal 31 Mar, 2026 Reviewers agreed at journal 30 Mar, 2026 Reviewers agreed at journal 30 Mar, 2026 Reviewers invited by journal 30 Mar, 2026 Editor assigned by journal 27 Mar, 2026 Submission checks completed at journal 27 Mar, 2026 First submitted to journal 27 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9210282","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":615152036,"identity":"89d39a20-b4cb-48e7-9e36-b7ed16956402","order_by":0,"name":"Tushar Sarkar","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAsklEQVRIiWNgGAWjYJACxgYGBjnilfNAtRiTriWxgWgt9vyHn32cUXMvfXv/GgOGHzXE2CKRZjxzw7Hi3Dk33hgw9hwjSguDMeMDtoTcGRJnDBh4iXEeD//xz4wP/iWkSwC1MP4lSgtDjjHjxraEBAn+HgNm4my5kVPMOLMvwXCGBFvBYRli/MLef3wzY8+3BHkJ/sMbH74hJsQQQCKB4QBJGhgY+EnVMApGwSgYBSMGAADDXTL38HkWwAAAAABJRU5ErkJggg==","orcid":"","institution":"The Assam Royal Global University","correspondingAuthor":true,"prefix":"","firstName":"Tushar","middleName":"","lastName":"Sarkar","suffix":""},{"id":615152037,"identity":"893e909d-7039-44ab-9194-efccf86dedef","order_by":1,"name":"Shobhit Pipil","email":"","orcid":"","institution":"The Assam Royal Global University","correspondingAuthor":false,"prefix":"","firstName":"Shobhit","middleName":"","lastName":"Pipil","suffix":""},{"id":615152038,"identity":"8b0160b3-0a82-44d5-838a-513234ce2577","order_by":2,"name":"Arghadeep Bose","email":"","orcid":"","institution":"The Assam Royal Global University","correspondingAuthor":false,"prefix":"","firstName":"Arghadeep","middleName":"","lastName":"Bose","suffix":""},{"id":615152039,"identity":"33c23af5-559c-41b3-8daf-bd0091e69d86","order_by":3,"name":"Teresa Sakrinsow","email":"","orcid":"","institution":"Vivekananda Kendra College of Teacher Education","correspondingAuthor":false,"prefix":"","firstName":"Teresa","middleName":"","lastName":"Sakrinsow","suffix":""}],"badges":[],"createdAt":"2026-03-24 10:08:39","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9210282/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9210282/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106043965,"identity":"fc37dcee-44ff-447d-b3cb-94ecd674b278","added_by":"auto","created_at":"2026-04-02 18:41:08","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":478941,"visible":true,"origin":"","legend":"\u003cp\u003eLocation of the study area and spatial distribution of schools in West Kameng district, Arunachal Pradesh\u003c/p\u003e","description":"","filename":"image1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9210282/v1/fc814175dce62abc5c45186e.jpeg"},{"id":106043966,"identity":"f360c097-18f6-4692-bcf3-76f270dd28e4","added_by":"auto","created_at":"2026-04-02 18:41:08","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1134622,"visible":true,"origin":"","legend":"\u003cp\u003eMethodological framework for terrain-constrained school accessibility and dropout analysis\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-9210282/v1/9e13f2c5c2ed3f14fb585785.png"},{"id":106093843,"identity":"3b01678d-ade0-4fed-b51a-2590ff93164d","added_by":"auto","created_at":"2026-04-03 11:39:30","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":785657,"visible":true,"origin":"","legend":"\u003cp\u003eSlope-derived resistance classification of West Kameng District, Arunachal Pradesh\u003c/p\u003e","description":"","filename":"image3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9210282/v1/e64c3ddfa477715f01d9e921.jpeg"},{"id":106043968,"identity":"eca0820e-91ad-4cec-a775-03fc946b2a93","added_by":"auto","created_at":"2026-04-02 18:41:08","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":306918,"visible":true,"origin":"","legend":"\u003cp\u003eLand use/land cover (ESA, 2021) and derived land-use resistance surface of West Kameng District, Arunachal Pradesh\u003c/p\u003e","description":"","filename":"image4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9210282/v1/5baa26575fa7000e73608f4b.jpeg"},{"id":106094419,"identity":"c1f6df32-cb46-425d-806c-154f01cca8cd","added_by":"auto","created_at":"2026-04-03 11:42:30","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":621159,"visible":true,"origin":"","legend":"\u003cp\u003eComposite terrain-induced cost surface of West Kameng District, Arunachal Pradesh\u003c/p\u003e","description":"","filename":"image5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9210282/v1/90896ad5d956b2e8243d9f2a.jpeg"},{"id":106093950,"identity":"72fcdb02-0361-4e72-a737-fd83c58da1b0","added_by":"auto","created_at":"2026-04-03 11:40:17","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":350239,"visible":true,"origin":"","legend":"\u003cp\u003eElevation gradient of accessibility (SAI) for primary schools in West Kameng District, Arunachal Pradesh\u003c/p\u003e","description":"","filename":"image6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9210282/v1/f1075ea0cf6bd69e766d0b4f.jpeg"},{"id":106043970,"identity":"33d30b4e-cd6b-4006-a636-6d150b24a9db","added_by":"auto","created_at":"2026-04-02 18:41:08","extension":"jpeg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":355909,"visible":true,"origin":"","legend":"\u003cp\u003eElevation gradient of accessibility (SAI) for middle schools in West Kameng District, Arunachal Pradesh\u003c/p\u003e","description":"","filename":"image7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9210282/v1/1fcc39fc47d16464b58bb1ee.jpeg"},{"id":106043971,"identity":"9013b9c5-16b5-45da-b5ff-ebe3b14f9fbf","added_by":"auto","created_at":"2026-04-02 18:41:08","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":341744,"visible":true,"origin":"","legend":"\u003cp\u003eElevation gradient of accessibility (SAI) for secondary and above schools in West Kameng District, Arunachal Pradesh\u003c/p\u003e","description":"","filename":"image8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9210282/v1/279e51a337f47c3059315fbe.jpeg"},{"id":106095879,"identity":"f3939d12-06e5-48ea-99e2-eb276da6238b","added_by":"auto","created_at":"2026-04-03 11:51:26","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6437445,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9210282/v1/a1326184-4f96-4999-953e-55dd86c2a519.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Analysing the relationship between terrain constrained school accessibility and dropout patterns in West Kameng District Arunachal Pradesh India","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eEducational access in mountainous regions is shaped by multiple structural and spatial factors. Together, institutional availability, settlement dispersion, transport infrastructure, and household socioeconomic conditions influence school participation in such a setting. However, the physical landscape imposes a foundational constraint. The terrain limits movement, regulates connectivity, and conditions the effective reach of educational institutions [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Elevation, slope, surface roughness, and dense vegetation increase the effort required to traverse space [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. For school-going children, this embodied difficulty often translates into irregular attendance. This often leads to early withdrawal from schooling [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. Educational accessibility in mountainous contexts must therefore be understood as materially grounded in terrain-conditioned mobility. However, dropout outcomes are shaped by multiple social, economic, and institutional factors alongside physical accessibility. This cannot be explained by institutional proximity alone.\u003c/p\u003e \u003cp\u003eExisting empirical research has largely examined the relationship between physical accessibility and schooling outcomes. Distance to school has been consistently identified as a critical determinant of enrolment and retention [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Evidence from Sub-Saharan Africa and South Asia shows that travel burdens affect schooling continuity across age groups [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. In India, regional studies likewise link remoteness and infrastructural deficits to gaps in educational participation [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. However, accessibility in most studies is measured through straight-line distance, travel time, or road connectivity. These measures capture proximity but not physical effort. Terrain is acknowledged but rarely operationalised as an independent explanatory factor. This limitation becomes more pronounced in mountainous environments where physical distance alone fails to represent movement difficulty. Studies in Himalayan regions have long documented the isolating effects of terrain on settlement systems and service delivery [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. More recent work in Arunachal Pradesh has also highlighted terrain as a barrier to educational access [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. However, quantitative modelling of terrain within formal accessibility frameworks remains limited.\u003c/p\u003e \u003cp\u003eAdvances in geographic information systems offer tools to address this gap [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Potential accessibility frames access as a function of spatial impedance rather than realised behaviour [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Cost-distance modelling incorporates terrain resistance through slope and land-cover surfaces [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Composite indices derived from such models have been widely used to evaluate access to health care and public services [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. The applications of such indices to schooling systems remain relatively rare. This is more evident in high-relief regions of South Asia. Arunachal Pradesh provides a critical empirical context in this regard. The state is characterised by extreme topographic variability, dispersed settlements, and limited transport penetration. The West Kameng district exemplifies similar conditions. Settlements are often distributed across steep valleys and forested uplands where daily mobility is physically demanding. Schools may exist within administrative proximity, yet their effective accessibility varies sharply across terrain gradients.\u003c/p\u003e \u003cp\u003eAgainst this backdrop, the present study examines terrain-conditioned accessibility within a GIS-based spatial framework. A school accessibility index (SAI) is constructed by integrating slope and land-use/land-cover resistance through cost-distance modelling. Accessibility conditions are evaluated across primary, middle, and secondary schooling stages. They are examined in relation to school-level dropout ratios. The analysis seeks to understand how terrain-induced accessibility varies spatially across the West Kameng district. It further explores whether these spatial accessibility patterns correspond with observed dropout patterns across school stages.\u003c/p\u003e"},{"header":"2 Theoretical Framework","content":"\u003cp\u003eThis study is situated within spatial accessibility theory. Accessibility is understood as a function of impedance rather than proximity [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. The conceptual foundation derives from the potential accessibility formulation proposed by Hansen (1959) [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. In this framework, accessibility reflects the resistance separating populations from opportunities. It represents structural possibility rather than realised behaviour. This perspective is particularly relevant in environments where travel data are unavailable. Accessibility can therefore be modelled as a property of spatial structure rather than individual mobility. However, the impedance is not uniform. It varies across physical landscapes. In mountainous regions, terrain constitutes a primary impedance field. Movement is structured by elevation gradients and slope morphology. Tobler (1993) [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e] demonstrated that travel velocity changes systematically with gradient. Surface conditions introduce additional resistance. Vegetation density reduces permeability. Forest cover slows movement. Rivers and streams interrupt travel pathways. Ray \u0026amp; Ebener (2008) [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e] reported that such barriers significantly alter accessibility patterns. Terrain impedance thus emerges as a composite function of topography and land cover.\u003c/p\u003e \u003cp\u003eCost-distance modelling operationalises this impedance structure within a spatial analytical framework [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. It moves beyond Euclidean proximity. Instead, it estimates cumulative resistance across continuous terrain surfaces. Each spatial unit contributes incrementally to travel effort. Movement pathways adjust in response to resistance. Accessibility is therefore expressed as least accumulated effort rather than geometric distance. Such modelling has been widely applied in service accessibility research, including health and infrastructure planning [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Its application to schooling accessibility, particularly in mountainous regions, remains comparatively limited. Within this analytical structure, the school accessibility index functions as a composite accessibility metric. It integrates slope and land cover resistance within a unified cost surface. The index captures spatial variation in terrain-conditioned accessibility across school locations. These accessibility differentials reflect structural mobility gradients. Such gradients provide a spatial context for examining how terrain-related accessibility conditions correspond with schooling outcomes such as dropout.\u003c/p\u003e"},{"header":"3 Study Area","content":"\u003cp\u003eWest Kameng District lies in the western part of Arunachal Pradesh in northeastern India. It forms a part of the Eastern Himalayan mountain system. The district shares international boundaries with Bhutan and the Tibet Autonomous Region of China. Physiographically, it is characterised by rugged relief and steep slopes [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. The valleys are deeply incised and spatially fragmented. The elevation varies sharply across short horizontal distances. This produces high terrain heterogeneity [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Dense forest cover further increases surface roughness. River gorges and unstable slopes increase movement difficulty. These geomorphic conditions impose strong constraints on mobility and spatial connectivity [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe settlement and service distribution in West Kameng reflect this terrain structure. The villages are small and spatially dispersed. Many are located along slopes and ridge corridors. Road penetration remains uneven across the district [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Foot-based mobility continues to play a significant role in everyday travel. Under such conditions, institutional proximity does not necessarily ensure functional access. Educational infrastructure operates within this terrain-constrained landscape. Students often traverse steep gradients and forested paths to reach schools [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Accessibility therefore varies sharply across short spatial distances. These characteristics make West Kameng an analytically appropriate setting for terrain-based accessibility modelling.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"4 Database","content":"\u003cp\u003eThis study adopts a GIS-based potential accessibility framework to assess terrain-induced constraints on access to schools. Accessibility is treated as a spatial property of the landscape. It is not interpreted as observing student travel behaviour. The analysis focuses on the relative physical difficulty imposed by existing terrain conditions. Terrain has been identified as an important constraint to educational access in Arunachal Pradesh [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. In this study, terrain is operationalised explicitly within the accessibility framework as a conditioning factor. Other determinants of dropout such as household socio-economic conditions, institutional quality, and cultural factors are acknowledged but are not directly modelled in this spatial analysis. The framework follows the concept of potential accessibility proposed by Hansen (1959) [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Here, accessibility is defined by spatial impedance. This approach is widely used in geography and spatial planning [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. This approach is suitable when reliable data on student residences and daily travel routes are unavailable [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eTerrain-induced accessibility is represented using slope and land-use/land-cover variables. Slope captures the dominant topographic constraint on walking. However, land cover represents surface-related barriers such as vegetation density and surface roughness. Together, these variables capture the primary physical properties of terrain relevant to pedestrian movement in mountainous environments. The school accessibility index (SAI) follows established practices in GIS-based accessibility modelling [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Cost-distance approaches have been used to represent spatial impedance through resistance surfaces [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Normalised composite indices have been applied to compare access to services such as schools and health facilities [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. The SAI builds on this tradition. It provides a relative and comparable measure of terrain-induced accessibility across schools and educational stages.\u003c/p\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e4.1 School location data\u003c/h2\u003e \u003cp\u003eSchool locations constitute the primary spatial input for the study. A total of 209 schools located in West Kameng district were identified from the Unified District Information System for Education Plus (UDISE+) Schools Geo-Portal. A ready-to-use GIS point layer is not available from any official sources. Therefore, school locations have been manually digitised using Google Earth Pro. Each school has been identified using high-resolution satellite imagery. Locations have been cross-verified with information available on the UDISE+ portal. Each digitised school has been assigned a unique school ID corresponding to the official UDISE code. Schools have been classified according to the highest-grade category offered for analytical standardisation. All primary schools have been treated as offering education up to Class V. All middle schools have been treated as offering education up to Class VIII. Schools offering education from Class IX onwards have been classified as secondary and above. This standard has been adopted solely to define school categories and associated buffer distances. This does not imply uniformity in the actual grade structure across schools.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e4.2 School attribute data\u003c/h2\u003e \u003cp\u003eSchool-level attribute data were compiled from Department of School Education, Government of Arunachal Pradesh. These data include total enrolment and total dropout. The data represent combined enrolment and dropout figures for the 2023-24 and 2024-25 academic years. The two years were combined to obtain a stable school-level dropout ratio. Combining consecutive academic years reduces year-specific fluctuations.\u003c/p\u003e \u003cp\u003eDropout has been operationalised as a school-level dropout ratio rather than an absolute count. The dropout ratio for each school was calculated following the procedure outlined by the Government of India (2023) [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] as follows:\u003c/p\u003e \u003cp\u003eDropout ratio =\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\:\\frac{Total\\:dropout\\:in\\:a\\:year}{Total\\:enrolment\\:at\\:the\\:begining\\:of\\:the\\:year}\\times\\:100\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003cp\u003eThis formulation standardises dropout across schools with different enrolment sizes. It enables comparisons across accessibility classes. Individual student trajectories have not been analysed. The attribute data were linked to the spatial school layer using the School ID. Schools lacking enrolment records were excluded from the analysis. Most missing data reflect administrative restructuring in the district. Several schools were merged with nearby institutions. Some were closed owing to persistently low enrolment. These changes are not systematically related to accessibility conditions. Among the 209 identified schools, 86 institutions (41.15%) were excluded for these reasons. The final analytical sample therefore includes 41 primary schools, 58 middle schools, and 24 secondary and above schools. This retained sample remains adequate for examining relative terrain-induced accessibility effects rather than absolute dropout levels.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Digital elevation model (DEM)\u003c/h2\u003e \u003cp\u003eTopographic data were obtained from the Copernicus Sentinel Digital Elevation Model (DEM). The DEM has a spatial resolution of 10 metres. It has been selected to represent rugged mountainous terrain. The resolution has been found to be compatible with other datasets used in the study. The DEM has been projected to WGS 84, UTM Zone 46 N, and has been clipped to the administrative boundary of West Kameng district.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Land use/land cover data\u003c/h2\u003e \u003cp\u003eThe land use and land cover data were obtained from the ESA WorldCover Version 2 (2021) dataset with a 10-meter spatial resolution [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. The dataset provides a consistent classification of surface cover types.\u003c/p\u003e \u003c/div\u003e"},{"header":"5 Methodology","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e5.1 Spatial harmonisation and preprocessing\u003c/h2\u003e \u003cp\u003eAll spatial datasets were projected to a common coordinate reference system, WGS 84, UTM Zone 46 N. The raster layers were aligned to a spatial resolution of 10 metres. Consistent snap raster and extent settings have been applied. This approach ensures pixel-level correspondence across raster layers.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e5.2 Slope derivation and terrain resistance modelling\u003c/h2\u003e \u003cp\u003eThe slope is derived from the DEM and represented in degrees. The slope angle provides a direct indicator of walking difficulty in mountainous terrain [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. The continuous slope surface was reclassified into five categories representing increasing levels of movement difficulty. The slope classification follows thresholds commonly used in land evaluation and geomorphological studies, as formalised in the FAO (1976) [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] land evaluation guidelines. These thresholds group slopes according to increasing physical limitations to human activity. The classes have been adapted to reflect pedestrian movement difficulty in the Himalayan terrain. Each slope class was assigned an ordinal resistance value ranging from 1 to 5. The resistance values represent the relative walking difficulty. They do not represent the calibrated physical cost.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003eSlope classes and resistance values used in terrain accessibility modelling\u003c/b\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSlope class\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSlope range (degrees)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTerrain interpretation\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eResistance value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVery gentle\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;3\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eEasy walking\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGentle\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3\u0026ndash;6\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSlight effort\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModerate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6\u0026ndash;10\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eModerate effort\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSteep\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10\u0026ndash;20\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDifficult walking\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVery steep\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u0026gt;\u0026thinsp;20\u0026deg;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSevere difficulty\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e5.3 Land cover resistance modelling\u003c/h2\u003e \u003cp\u003eThe land cover resistance was modelled using the WorldCover dataset [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. Each land-cover class was evaluated for pedestrian movement difficulty. The classes have been reclassified into resistance values ranging from 1\u0026ndash;5 [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. The same ordinal scale used for slope resistance was applied to maintain conceptual consistency. Dense forest and water bodies were assigned the maximum resistance values. These classes represent severe constraints on movement. This approach avoids introducing absolute barriers while retaining realistic terrain impedance.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003eLand-use/land-cover classes and resistance values used in accessibility modelling\u003c/b\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLand-cover class\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDescription\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eResistance value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAccessibility implication\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBuilt-up\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSettlements, roads\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eVery easy movement\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCropland / Grassland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAgricultural land\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eEasy movement\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShrubland\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSecondary vegetation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eModerate obstruction\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTree cover\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOpen or mixed forest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDifficult movement\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMoss and lichen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAlpine forest, sparse ground cover\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDifficult movement due to cold, moisture, uneven surface\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDense forest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eClosed canopy forest\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eVery difficult movement\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWater bodies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRivers, lakes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSevere constraint\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSnow and ice\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePermanent snow, glaciers\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSevere constraint; often impassable\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e5.4 Construction of the composite cost surface\u003c/h2\u003e \u003cp\u003eSlope resistance and land cover resistance have been integrated to create a composite cost surface [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Both components have been assigned equal weights. The composite cost was calculated as the linear aggregation of the two resistance values. Equal weighting has been adopted deliberately. Resistance values are ordinal indicators of relative movement difficulty. They do not represent the calibrated physical cost. Empirical data quantifying the differential contributions of slope and land cover are not available. Equal weighting therefore provides a conservative modelling choice. This approach avoids introducing subjective bias. The composite surface represents the cumulative terrain-induced constraint rather than the absolute travel cost. Sensitivity testing of alternative weighting schemes has not been conducted because of the absence of calibrated pedestrian movement data for the district.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e5.5 Cost-distance modelling\u003c/h2\u003e \u003cp\u003eCost-distance modelling has been employed to estimate potential physical accessibility under terrain-induced constraints [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Unlike Euclidean distance, cost-distance analysis accounts for spatial heterogeneity in movement difficulty. It incorporates resistance values derived from terrain and surface conditions. In mountainous regions, the straight-line distance poorly represents the actual movement effort. Therefore, cost-distance modelling provides a more realistic approximation of terrain-conditioned accessibility [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Using the composite cost surface, cost‒distance analysis was performed with schools treated as destination points. The analysis calculates the least cumulative cost required to reach the nearest school from each raster cell. The resulting raster represents potential physical accessibility rather than observed travel behaviour.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e5.6 Delineation of school-centred accessibility buffers\u003c/h2\u003e \u003cp\u003eThe student residence data are not available from official sources. School-centred buffers have therefore been used to approximate potential catchment areas. Buffer distance is defined according to the educational stage and the expected daily mobility of students. A 1 km, 2 km, and 3 km buffer has been applied to primary, middle, and secondary and above schools, respectively [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. These buffers represent potential catchment extents. They approximate potential accessibility zones rather than actual travel routes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e5.7 Extraction of school-level accessibility measures\u003c/h2\u003e \u003cp\u003eZonal statistics were computed in ArcGIS Pro using Zonal Statistics as a table tool. School buffer polygons were used as zones. The cost distance raster served as the input value raster. For each buffer, the mean cost distance value is extracted. This value represents the average terrain-induced movement effort within the potential catchment of each school. The extracted values have been linked back to the school attribute table using the School ID.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e5.8 Construction of the School Accessibility Index (SAI)\u003c/h2\u003e \u003cp\u003eThe school accessibility index (SAI) was calculated using min-max normalisation of the mean cost-distance values. This follows established GIS-based accessibility indices in educational geography [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. It has been performed separately for primary, middle, and secondary and above schools. This accounts for differences in buffer size and accessibility context across educational stages.\u003c/p\u003e \u003cp\u003eThe school accessibility index (SAI) is calculated as follows:\u003c/p\u003e \u003cp\u003eSAI\u003csub\u003ei\u003c/sub\u003e =\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{{C}_{i}-{C}_{min}}{{C}_{max}-{C}_{min}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e \u003cp\u003eHere, C\u003csub\u003ei\u003c/sub\u003e represents the mean cost-distance value for school i.\u003c/p\u003e \u003cp\u003eCₘ\u003csub\u003ei\u003c/sub\u003eₙ and Cₘₐₓ represent the minimum and maximum values within the same school category.\u003c/p\u003e \u003cp\u003eLower SAI values indicate greater accessibility. Higher values indicate greater terrain-induced constraints.\u003c/p\u003e \u003cp\u003eThe SAI values are grouped into four accessibility zones. The classification was carried out separately for each school category.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003eSchool Accessibility Index (SAI) classification scheme\u003c/b\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSAI range\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAccessibility class\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eInterpretation\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e0.00\u0026ndash;0.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh accessibility\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMinimal terrain-induced constraint\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e0.26\u0026ndash;0.50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eModerate accessibility\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLow to moderate terrain constraint\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e0.51\u0026ndash;0.75\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLow accessibility\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHigh terrain-induced constraint\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e0.76\u0026ndash;1.00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVery low accessibility\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSevere terrain-induced constraint\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e5.9 Statistical analysis of accessibility and dropout\u003c/h2\u003e \u003cp\u003eThe relationship between terrain-induced accessibility and school dropout has been examined using nonparametric statistical methods. These methods are appropriate because of their small sample sizes, nonnormal distributions, and the presence of tied values [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Spearman\u0026rsquo;s rank correlation was initially considered to assess the monotonic association between the school accessibility index (SAI) and the school-level dropout ratio. The SAI is treated as an ordinal variable. However, the correlation could not be reliably estimated because of extensive tied ranks in both accessibility classes and dropout ratios. Therefore, Kruskal-Wallis H tests were applied to examine differences in dropout ratios across SAI classes. The tests were conducted separately for primary, middle, and secondary schools and above.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"6 Results and Discussion","content":"\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\n \u003ch2\u003e6.1 Terrain resistance structure of West Kameng district\u003c/h2\u003e\n \u003cp\u003eSlope and land-use resistance together produce a highly constrained mobility surface across the West Kameng district. Steep gradients dominate the landscape (Fig. \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e; Table \u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The highest slope resistance class (very steep) alone accounts for 76.12% of the total geographical area of the district (Table \u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). In contrast, very gentle and gentle slopes together form only 2.41% of the district\u0026rsquo;s total area. This indicates that most settlements are embedded within physically demanding terrain. Daily movement therefore requires greater time and effort. This suggests that school travel becomes physically burdensome in such environments. Ives \u0026amp; Messerli (2003) [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e] documented a similar terrain‒mobility constraint across Himalayan mountain systems. They noted steep relief fragment settlement connectivity and limited service accessibility in mountainous regions.\u003c/p\u003e\n \u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u003cstrong\u003eAreal distribution of slope resistance classes\u003c/strong\u003e\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eSlope resistance value\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eSlope class\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003eArea (sq. km)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003e% Area\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eVery gentle\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e62.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0.84\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eGentle\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e116.99\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e1.57\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eModerate\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e252.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e3.41\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eSteep\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e1339.69\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e18.05\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eVery steep\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e5649.97\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e76.12\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003eLand cover resistance deepens this constraint. A large share of the district is covered by high-resistance surfaces. These mainly include tree cover (open and mixed forest), alpine forest, and areas with sparse ground cover. Together, they account for 91.56% of the district\u0026rsquo;s total area (Fig. \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e; Table \u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). In contrast, low-resistance surfaces such as built-up (0.16%) and cropland (8%) land remain spatially fragmented. These pockets rarely form continuous access corridors. The composite terrain classification reflects this cumulative resistance. Nearly 87.27% of the district falls within the severe terrain difficulty zone (Table\u0026nbsp;6). In contrast, very easy and easy terrains together account for less than 2%. Accessibility is therefore structurally valley bound and ridge constrained. Institutional reach becomes uneven. The terrain therefore forms the physical context within which spatial differences in school accessibility emerge. Similar cumulative resistance effects have been reported in accessibility modelling across South Asian and Southeast Asian Mountain regions [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. In such terrains, slope and surface impedance operate together to constrain movement.\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eTable 5\u003c/strong\u003e \u003cstrong\u003eAreal distribution of land-use/land-cover resistance classes\u003c/strong\u003e\u003c/p\u003e\n \u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 191px;\"\u003e\n \u003cp\u003eLULC resistance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 187px;\"\u003e\n \u003cp\u003eArea (sq. km)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 181px;\"\u003e\n \u003cp\u003e% Area\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 191px;\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 187px;\"\u003e\n \u003cp\u003e11.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 181px;\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 191px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 187px;\"\u003e\n \u003cp\u003e593.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 181px;\"\u003e\n \u003cp\u003e8.00\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 191px;\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 187px;\"\u003e\n \u003cp\u003e0.0086\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 181px;\"\u003e\n \u003cp\u003e0.0001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 191px;\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 187px;\"\u003e\n \u003cp\u003e6795.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 181px;\"\u003e\n \u003cp\u003e91.56\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 191px;\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 187px;\"\u003e\n \u003cp\u003e21.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 181px;\"\u003e\n \u003cp\u003e0.29\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eTable 6\u0026nbsp;\u003c/strong\u003eA\u003cstrong\u003ereal distribution of composite terrain resistance zones\u003c/strong\u003e\u003c/p\u003e\n \u003cdiv align=\"\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003eZone\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 171px;\"\u003e\n \u003cp\u003eArea (sq. km)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 166px;\"\u003e\n \u003cp\u003e% Area\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003eVery Easy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 171px;\"\u003e\n \u003cp\u003e21.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 166px;\"\u003e\n \u003cp\u003e0.29\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003eEasy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 171px;\"\u003e\n \u003cp\u003e111.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 166px;\"\u003e\n \u003cp\u003e1.50\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003eModerate\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 171px;\"\u003e\n \u003cp\u003e248.72\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 166px;\"\u003e\n \u003cp\u003e3.35\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003eDifficult\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 171px;\"\u003e\n \u003cp\u003e563.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 166px;\"\u003e\n \u003cp\u003e7.59\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003eSevere\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 171px;\"\u003e\n \u003cp\u003e6477.03\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 166px;\"\u003e\n \u003cp\u003e87.27\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\n \u003ch2\u003e6.2 Composite terrain cost surface of West Kameng district\u003c/h2\u003e\n \u003cp\u003eThe composite terrain cost surface integrates slope steepness and land cover resistance into a continuous representation of cumulative movement difficulty. High-cost terrain dominates almost the entire district (Fig. \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This pattern is particularly evident across the upland and dissected interior regions. These areas coincide with high mountain ridges and densely forested slopes. The terrain is structurally dissected. Movement across such landscapes requires considerable physical effort. Travel time may also increase under such terrain conditions. Seasonal conditions such as landslides or surface wetness could further intensify these constraints. This ridge-concentrated resistance pattern aligns with the anisotropic cost-distance modelling approaches applied in mountainous regions. Studies from high-relief landscapes, including parts of South Asia and Southeast Asia, have shown that slope acts as a directional impedance. It alters effective movement pathways and increases cumulative travel effort. Similar modelling applications have demonstrated how ridge systems restructure accessibility surfaces in terrain-dominated environments [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e].\u003c/p\u003e\n \u003cp\u003eLow-cost corridors occur only in fragmented pockets. These belts are largely confined to valley floors and river-aligned settlement belts (Fig. \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The Kameng River valley and its tributary systems form the corridors with the most visible accessibility. The terrain in these zones appears relatively navigable. Settlement clustering and road presence are also more evident. Schools located within such corridors are likely to have relatively wider functional catchments. In contrast, institutions situated along ridge slopes or forested uplands remain embedded within high cumulative cost zones. Their accessibility reach is likely to remain spatially restricted. The composite surface therefore suggests that terrain-induced mobility gradients structure spatial differences in potential school accessibility across the district.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\n \u003ch2\u003e6.3 Spatial inequality in accessibility\u003c/h2\u003e\n \u003cp\u003eAccessibility inequality across West Kameng District follows a clear elevation gradient. Primary schools remain heavily concentrated below 2000 metres (Fig. \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e; Table \u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e7\u003c/span\u003e). The 1000\u0026ndash;2000 metre belt alone accounts for 29 (70.73%) schools out of the 41 total primary schools. This finding indicates that the lower-altitude valley system forms the main institutional zone at the foundational stage. Within this belt, 51.72 percent fall within moderate-accessibility zones, and 27.59 percent fall within high-accessibility zones (Table \u003cspan refid=\"Tab6\" class=\"InternalRef\"\u003e7\u003c/span\u003e). However, another 20.69 percent of the cases occur in low-accessibility terrain. However, it is worth noting that no primary school occurs within the very low accessibility category at any elevation. This pattern indicates that foundational schooling has largely been positioned within terrain environments that remain physically manageable for younger children.\u003c/p\u003e\n \u003cp\u003eAccessibility inequality at the middle stage broadly follows the primary pattern. However, it begins to expand into more difficult terrain. The 1000\u0026ndash;2000 meter belt also recorded a high concentration of middle schools, which closely mirrors the distribution of primary schools within this elevation range. This belt accounted for 39 schools (67.24%) out of the total 58\u003c/p\u003e\n \u003ctable float=\"Yes\" id=\"Tab6\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 7\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u003cstrong\u003eElevation-wise distribution of primary schools across accessibility (SAI) classes\u003c/strong\u003e\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\n \u003cp\u003eElevation (Meter)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\n \u003cp\u003eSAI class\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eHigh accessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003eModerate\u003c/p\u003e\n \u003cp\u003eaccessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003eLow\u003c/p\u003e\n \u003cp\u003eaccessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c5\"\u003e\n \u003cp\u003eVery low accessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;1000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003cp\u003e(20.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003cp\u003e(40.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003cp\u003e(40.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e1000\u0026ndash;2000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003cp\u003e(27.59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003cp\u003e(51.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003cp\u003e(20.69)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e29\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e2001\u0026ndash;3000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003cp\u003e(14.29)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003cp\u003e(57.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003cp\u003e(28.57)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e3001\u0026ndash;4000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e\u0026gt;\u0026thinsp;4000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003cp\u003e(24.29)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e21\u003c/p\u003e\n \u003cp\u003e(51.22)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003cp\u003e(24.39)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e41\u003c/p\u003e\n \u003cp\u003e(100)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"6\"\u003e\u003cstrong\u003eNote\u003c/strong\u003e: Figures in parentheses are percentages\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n \u003ctable float=\"Yes\" id=\"Tab7\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 8\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u003cstrong\u003eElevation- distribution of middle schools across accessibility (SAI) classes\u003c/strong\u003e\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\n \u003cp\u003eElevation (Meter)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\n \u003cp\u003eSAI class\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eHigh accessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003eModerate accessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003eLow accessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c5\"\u003e\n \u003cp\u003eVery low accessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;1000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003cp\u003e(42.86)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003cp\u003e(42.86)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003cp\u003e(14.29)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e1000\u0026ndash;2000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003cp\u003e(35.90)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003cp\u003e(48.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003cp\u003e(10.26)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003cp\u003e(5.13)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e39\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e2001\u0026ndash;3000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003cp\u003e(25.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003cp\u003e(41.67)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003cp\u003e(25.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003cp\u003e(8.33)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e3001\u0026ndash;4000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e\u0026gt;\u0026thinsp;4000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003cp\u003e(34.48)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e27\u003c/p\u003e\n \u003cp\u003e(46.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003cp\u003e(13.79)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003cp\u003e(5.17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e58\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"6\"\u003e\u003cstrong\u003eNote\u003c/strong\u003e: Figures in parentheses are percentages\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n \u003cp\u003emiddle schools in the district (Fig. \u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e; Table \u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e8\u003c/span\u003e). This confirms the continued concentration within lower-altitude settlement corridors. Within this belt, 48.72% of the schools fall in moderate-accessibility zones, and 35.90% fall in high-accessibility zones (Table \u003cspan refid=\"Tab7\" class=\"InternalRef\"\u003e8\u003c/span\u003e). However, 15.39% of cases occur in low- and very low-accessibility terrains. Terrain exposure therefore widens beyond the foundational stage. The shift becomes clearer above 2000 metres. In the 2001\u0026ndash;3000 m belt, 25.00% of the transects fall in high-accessibility zones. However, 33.33% of the schools operate under low- and very-low-accessibility conditions. Compared with primary schools, middle institutions show greater terrain exposure.\u003c/p\u003e\n \u003cp\u003eThis gradient becomes more pronounced at the secondary level and above. In contrast to the primary and middle stages, the concentration of schools declines in this tier. The 1000\u0026ndash;2000 m belt contains 14 out of the 24 secondary schools and above, accounting for 58.33% of the total (Fig. \u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e; Table \u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e9\u003c/span\u003e). This indicates partial expansion beyond the core valley corridors. Within this belt, 64.29% of the area remains in high-accessibility terrain (Table \u003cspan refid=\"Tab8\" class=\"InternalRef\"\u003e9\u003c/span\u003e). Conditions become more challenging beyond 2000 metres. In the 2001\u0026ndash;3000 m belt, only 12.50% of the transects fall in high-accessibility zones. However, 50.00% of the cases occur in low- and very low-accessibility environments. A clear progression therefore emerges. Primary schools remain valley-buffered. The middle schools have moderate terrain exposure. Secondary institutions and higher-level institutions face the greatest constraints. Accessibility inequality therefore becomes more pronounced with educational progression and elevation.\u003c/p\u003e\n \u003ctable float=\"Yes\" id=\"Tab8\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 9\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u003cstrong\u003eElevation wise distribution of secondary and above school across accessibility (SAI) classes\u003c/strong\u003e\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\n \u003cp\u003eElevation (Meter)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e\n \u003cp\u003eSAI class\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eHigh accessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003eModerate accessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003eLow accessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c5\"\u003e\n \u003cp\u003eVery low accessibility\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;1000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003cp\u003e(50.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003cp\u003e(50.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e1000\u0026ndash;2000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003cp\u003e(64.29)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003cp\u003e(28.57)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003cp\u003e(7.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e2001\u0026ndash;3000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003cp\u003e(12.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003cp\u003e(37.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003cp\u003e(12.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003cp\u003e(37.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e3001\u0026ndash;4000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003e\u0026gt;\u0026thinsp;4000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003cp\u003e(0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003cp\u003e(45.83)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003cp\u003e(29.17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003cp\u003e(8.33)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003cp\u003e(16.67)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\n \u003cp\u003e24\u003c/p\u003e\n \u003cp\u003e(100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"6\"\u003e\u003cstrong\u003eNote\u003c/strong\u003e: Figures in parentheses are percentages\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec23\" class=\"Section2\"\u003e\n \u003ch2\u003e6.4 Dropout patterns across accessibility zones in West Kameng district\u003c/h2\u003e\n \u003cp\u003eDropout patterns vary across accessibility classes. However, the strength of the association differs by schooling stage. At the primary level, the pattern is irregular. The mean dropout rate is highest in high-accessibility zones, at 11.09% (Table \u003cspan refid=\"Tab9\" class=\"InternalRef\"\u003e10\u003c/span\u003e). It decreases sharply in moderate accessibility areas to 3.72%. It rises again in low-accessibility terrain to 9.65%. No school falls within the very low accessibility class. This absence of a consistent gradient suggests that terrain constraint does not appear to strongly structure early-stage withdrawal from schooling. At this stage, household conditions or schooling readiness may play a greater role than travel burden alone. Huisman \u0026amp; Smits (2009) [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] reported a similar pattern in mountainous regions. Their findings suggest that household vulnerability often has a stronger effect than travel distance in shaping early-grade retention and dropout.\u003c/p\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n \u003ctable float=\"Yes\" id=\"Tab9\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 10\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u003cstrong\u003eDistribution of primary schools and mean dropout ratio across accessibility (SAI) classes\u003c/strong\u003e\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eSAI class\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eNo. of schools\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003ePercentage of school\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003eMean dropout ratio (%)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eHigh accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e24.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003e11.09\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eModerate accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e51.22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003e3.72\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eLow accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e24.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003e9.65\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eVery low accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003e-\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eAll primary schools\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e100.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003e6.96\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003eThe influence of the terrain constraint becomes more visible beyond the primary level. At the middle stage, the dropout rate increases steadily as accessibility decreases. It remains lowest in moderate-accessibility zones at 3.24% (Table \u003cspan refid=\"Tab10\" class=\"InternalRef\"\u003e11\u003c/span\u003e). It increases in high-accessibility areas to 5.42%. It increases further in low-accessibility terrain to 7.73%. The highest dropout rate of 9.31% occurs in very low-accessibility environments (Table \u003cspan refid=\"Tab10\" class=\"InternalRef\"\u003e11\u003c/span\u003e). This progression suggests that terrain-related travel effort may become more relevant during the transition from primary to middle schooling. Such distance-related discontinuation at post primary stages has also been documented across rural schooling systems in developing regions [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. The secondary and above stage shows a moderated but related pattern. Dropout remains lowest in high-accessibility zones at 3.44% (Table \u003cspan refid=\"Tab11\" class=\"InternalRef\"\u003e12\u003c/span\u003e). It increases in moderate- and low-accessibility environments to 9.17% and 8.44%, respectively. Very low-accessibility schools record a dropout ratio of 5.35 percent (Table \u003cspan refid=\"Tab11\" class=\"InternalRef\"\u003e12\u003c/span\u003e). Despite this variation, higher dropout levels remain concentrated outside the high-accessibility class. This partial gradient reflects findings from South Asian hill regions, where continuation beyond elementary schooling becomes mobility sensitive rather than proximity dependent [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e].\u003c/p\u003e\n \u003cp\u003eTaken together, terrain accessibility does not produce a uniform dropout response. Its role appears weak at the primary stage. A clearer descriptive gradient emerges during middle schooling and remains partially visible at the secondary and above level. As students progress to higher grades, commuting routes often become longer. Daily travel may also become more physically demanding. In such settings, terrain-related travel burden may interact with household and institutional conditions that influence school continuation. Terrain therefore does not directly determine dropout. Instead, it forms part of the broader spatial context within which withdrawal from schooling may occur.\u003c/p\u003e\n \u003ctable float=\"Yes\" id=\"Tab10\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 11\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u003cstrong\u003eDistribution of middle schools and mean dropout ratio across accessibility (SAI) classes\u003c/strong\u003e\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eSAI class\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eNo. of schools\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003ePercentage of school\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003eMean dropout ratio (%)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eHigh accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e34.48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e5.42\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eModerate accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e46.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e3.24\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eLow accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e13.79\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e7.73\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eVery low accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e5.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e9.31\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eAll middle schools\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e100.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e4.93\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n \u003ctable float=\"Yes\" id=\"Tab11\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 12\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u003cstrong\u003eDistribution of secondary and above schools and mean dropout ratio across accessibility (SAI) classes\u003c/strong\u003e\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eSAI class\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eNo. of schools\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003ePercentage of school\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c4\"\u003e\n \u003cp\u003eMean dropout ratio (%)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eHigh accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e45.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e3.44\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eModerate accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e29.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e9.17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eLow accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e8.33\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e8.44\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eVery low accessibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e16.67\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e5.35\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eAll secondary and above schools\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e100.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\n \u003cp\u003e5.84\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec24\" class=\"Section2\"\u003e\n \u003ch2\u003e6.5 Statistical relationship between accessibility and dropout\u003c/h2\u003e\n \u003cp\u003eNonparametric tests have been applied to examine whether dropout ratios vary significantly across accessibility classes. The Kruskal‒Wallis results show limited statistical differentiation. The H values of primary, middle and secondary and above schools were 0.457 (p\u0026thinsp;=\u0026thinsp;0.796), 2.611 (p\u0026thinsp;=\u0026thinsp;0.456) and 0.698 (p\u0026thinsp;=\u0026thinsp;0.874), respectively (Table \u003cspan refid=\"Tab12\" class=\"InternalRef\"\u003e13\u003c/span\u003e). In all the cases, the p values remain above the 0.05 threshold. Therefore, dropout variation across SAI classes is not statistically significant within the present dataset. However, descriptive patterns reveal a more nuanced gradient pattern. Dropout ratios generally rise as accessibility declines, particularly beyond the primary stage. Middle schools show greater dropout rates within low- and very-low-accessibility zones. Secondary and above schools also recorded comparatively higher dropout levels outside high-accessibility classes. These descriptive patterns suggest that terrain-related accessibility may operate as a contextual constraint rather than a statistically isolated predictor. The absence of statistical significance may reflect structural data limitations. Extreme accessibility categories contain fewer schools. This reduces the statistical power of nonparametric tests. The analysis is also conducted at the school level rather than at the level of individual student residences. Consequently, the terrain conditions experienced by individual students may not be fully captured by school-centred accessibility measures.\u003c/p\u003e\n \u003cp\u003eTerrain accessibility may therefore operate as a conditioning factor rather than a direct causal determinant. Its influence is likely to interact with institutional reach, travel fatigue, and stage-specific mobility demands. Accessibility inequality nevertheless remains analytically meaningful. In such terrain-dominated environments, spatial accessibility differences provide an important contextual dimension for understanding variation in schooling continuation.\u003c/p\u003e\n \u003ctable float=\"Yes\" id=\"Tab12\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 13\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u003cstrong\u003eKruskal\u0026ndash;Wallis test results for differences in dropout ratios across accessibility\u003c/strong\u003e (SAI) classes by school stage\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eStage of School\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c2\"\u003e\n \u003cp\u003eH Statistics\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\" colname=\"c3\"\u003e\n \u003cp\u003ep-value\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003ePrimary\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e0.457\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0.796\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eMiddle\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e2.611\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0.456\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" colname=\"c1\"\u003e\n \u003cp\u003eSecondary \u0026amp; above\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\n \u003cp\u003e0.698\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\n \u003cp\u003e0.874\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003c/div\u003e"},{"header":"7 Conclusions","content":"\u003cp\u003eThis study examined terrain-constrained physical accessibility to schools in West Kameng district using a GIS-based school accessibility index. It also explored how accessibility conditions are related to school-level pattern of dropout across educational stages. The findings indicate that terrain forms a structurally embedded constraint on educational access. Steep slopes dominate the district surface. Dense forest cover further intensifies movement difficulty. Severe terrain difficulty characterises most geographical areas. Accessibility therefore remains concentrated within valley corridors. Ridge and upland locations experience restricted mobility conditions.\u003c/p\u003e \u003cp\u003eThe school distribution reflects this terrain structure. Primary institutions remain largely valley-buffered. Most are located within moderate-accessibility environments. Terrain exposure increases during the middle and secondary stages. Higher-level schools extend into more physically demanding landscapes. Dropout patterns show a differentiated relationship. The influence of the terrain appears weak at the primary stage. A clearer descriptive gradient emerges during middle schooling. The pattern continues at the secondary and above level, although less uniformly. As students progress to a higher level of education, commuting routes often become longer. Daily travel may therefore become increasingly demanding under steep terrain conditions. The accessibility burden appears to accumulate with educational progression. The statistical tests do not reveal significant variation in dropout ratios across accessibility classes. However, descriptive patterns suggest that higher dropout levels are more frequently observed outside high-accessibility zones. This is more evident particularly beyond the primary stage of schooling. These findings indicate that terrain accessibility does not operate as an isolated determinant of dropout. Instead, it forms part of the broader spatial context within which schooling continuation takes place. In mountainous environments, mobility constraints may interact with institutional reach, household vulnerability, and travel fatigue. Terrain therefore acts as a conditioning environment that may amplify existing educational disadvantages rather than directly producing dropout.\u003c/p\u003e \u003cp\u003eThis study contributes by modelling terrain-conditioned accessibility within a GIS-based framework. The school accessibility index provides a stage-sensitive measure of accessibility differences across schools in mountainous terrain. The results suggest that accessibility inequalities become more visible beyond foundational schooling. These findings carry policy implications. Educational expansion in high-relief regions must consider terrain-conditioned mobility rather than administrative proximity alone. Residential schooling and transport support may therefore be important for sustaining post-primary education. Future research may incorporate household location data and actual student travel routes. Integrating socio-economic and institutional variables would further clarify how terrain accessibility interacts with broader determinants of school participation.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eEthics approval and consent to participate\u003c/h2\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eConsent for publication\u003c/strong\u003e \u003cp\u003eNot applicable.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eCompeting interests\u003c/strong\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis research did not receive any funding from any individual or organization.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eFirst and corresponding author conceptualized the study and prepared the original draft of the manuscript. First co-author prepared the maps using GIS and provided methodological support. Second co-author contributed to the methodological framework and supported discussion. 3rd co-author assisted in data collection and compilation.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets used in this study are publicly available. School-level data were obtained from the [UDISE+](https:/schoolgis.nic.in) portal. Spatial datasets include the [Copernicus Digital Elevation Model (DEM)](https:/dataspace.copernicus.eu/explore-data/data-collections/copernicus-contributing-missions/collections-description/COP-DEM) and [ESA WorldCover](https:/esa-worldcover.org/en/data-access) (2021). Processed data may be made available from the corresponding author upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbdul Yamin NAA, Basaruddin KS, Abu Bakar S, Salleh AF, Som M, Yazid MH, H., Hoang TD. (2022). Quantification of gait stability during incline and decline walking: The responses of required coefficient of friction and dynamic postural index. Journal of Healthcare Engineering, 2022(1), 7716821. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1155/2022/7716821\u003c/span\u003e\u003cspan address=\"10.1155/2022/7716821\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlam S. (2024, January 10). The obstacles to Arunachal's education system. The Arunachal Times. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://arunachaltimes.in/index.php/2024/01/10/the-obstacles-to-arunachals-education-system/\u003c/span\u003e\u003cspan address=\"https://arunachaltimes.in/index.php/2024/01/10/the-obstacles-to-arunachals-education-system/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (Accessed 20 February 2026).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAl-Sabbagh TA. GIS location-allocation models in improving accessibility to primary schools in Mansura. city-Egypt Geoj. 2022;87(2):1009\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10708-020-10290-5\u003c/span\u003e\u003cspan address=\"10.1007/s10708-020-10290-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBiswas G, Sharma MG. Impact of geographical factors on education in Papum Pare District of Arunachal Pradesh. Res J Educational Sci. 2018;6(2):1\u0026ndash;5. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://oldror.lbp.world/UploadedData/6740.pdf\u003c/span\u003e\u003cspan address=\"https://oldror.lbp.world/UploadedData/6740.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCentral Ground Water Board. (2013). Ground water information booklet: West Kameng District, Arunachal Pradesh. Ministry of Water Resources, Government of India. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.cgwb.gov.in/old_website/District_Profile/Arunachal/West%20Kameng.pdf\u003c/span\u003e\u003cspan address=\"https://www.cgwb.gov.in/old_website/District_Profile/Arunachal/West%20Kameng.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (Accessed 10 February 2026).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConover WJ. Practical nonparametric statistics. 3rd ed. Wiley; 1999.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCopernicus Programme. (2021). Copernicus digital elevation model (DEM) 10 m. European Space Agency. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5270/ESA-c5d3d65\u003c/span\u003e\u003cspan address=\"10.5270/ESA-c5d3d65\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDamurski Ł. Neighbourhood cohesion and territorial cohesion: in search for conceptual integrity. GeoJournal. 2022;87(6):4635\u0026ndash;51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10708-021-10523-1\u003c/span\u003e\u003cspan address=\"10.1007/s10708-021-10523-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDevelopment, Commissioner, Government of India. Ministry of Micro, Small \u0026amp; Medium Enterprises,. (2014). Brief industrial profile of West Kameng District. Government of India. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://dcmsme.gov.in/dips/DIP_West-Kameng-Dist.pdf\u003c/span\u003e\u003cspan address=\"https://dcmsme.gov.in/dips/DIP_West-Kameng-Dist.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (Accessed 7 February 2026).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDruga M, Min\u0026aacute;r J. Cost distance and potential accessibility as alternative spatial approximators of human influence in LUCC modelling. Land Use Policy. 2023;132:106840. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.landusepol.2023.106840\u003c/span\u003e\u003cspan address=\"10.1016/j.landusepol.2023.106840\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFilmer D. If you build it, will they come? School availability and school enrolment in 21 poor countries. J Dev Stud. 2007;43(5):901\u0026ndash;28. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00220380701384598\u003c/span\u003e\u003cspan address=\"10.1080/00220380701384598\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFood and Agriculture Organization of the United Nations. (1976). A framework for land evaluation (FAO Soils Bulletin No. 32). FAO. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.fao.org/4/x5310e/x5310e00.htm\u003c/span\u003e\u003cspan address=\"https://www.fao.org/4/x5310e/x5310e00.htm\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (Accessed 12 October 2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGagnon A. (2022). Isochrone-based catchment areas for educational planning. International Institute for Educational Planning (IIEP), UNESCO. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.iiep.unesco.org/en/publication/isochrone-based-catchment-areas-educational-planning\u003c/span\u003e\u003cspan address=\"https://www.iiep.unesco.org/en/publication/isochrone-based-catchment-areas-educational-planning\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (Accessed 1 January 2026).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeurs KT, van Wee B. Accessibility evaluation of land-use and transport strategies: Review and research directions. J Transp Geogr. 2004;12(2):127\u0026ndash;40. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jtrangeo.2003.10.005\u003c/span\u003e\u003cspan address=\"10.1016/j.jtrangeo.2003.10.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGlewwe P, Jacoby HG. An economic analysis of delayed primary school enrollment in a low-income country: The role of early childhood nutrition. Rev Econ Stat. 1995;77(1):156\u0026ndash;69. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.jstor.org/stable/2110001\u003c/span\u003e\u003cspan address=\"https://www.jstor.org/stable/2110001\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGovernment of India, Ministry of Education, Department of School Education and Literacy. (2023). School dropout rate: Concept, formula and procedure. Government of India. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://educationforallinindia.com/wp-content/uploads/2023/01/Drop-out-rate-formula.pdf\u003c/span\u003e\u003cspan address=\"https://educationforallinindia.com/wp-content/uploads/2023/01/Drop-out-rate-formula.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (Accessed 13 October 2026).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGovinda R, Bandyopadhyay M. (2008). Access to elementary education in India: Country analytical review. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://eric.ed.gov/?id=ED508807\u003c/span\u003e\u003cspan address=\"https://eric.ed.gov/?id=ED508807\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrogan A. (2009). Creating a spatial analysis model for generating composite cost surfaces to depict cross country mobility in natural terrain. In ASPRS/MAPPS Fall Conference. Texas.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHansen WG. How accessibility shapes land use. J Am Inst Planners. 1959;25(2):73\u0026ndash;6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01944365908978307\u003c/span\u003e\u003cspan address=\"10.1080/01944365908978307\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHengl T, Reuter HI. Geomorphometry: A brief guide. In: Hengl T, Reuter HI, editors. Geomorphometry: Concepts, software, applications. Elsevier; 2009. pp. 1\u0026ndash;33.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuisman J, Smits J. Effects of household- and district-level factors on primary school enrolment in 30 developing countries. World Dev. 2009;37(1):179\u0026ndash;93. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.worlddev.2008.01.007\u003c/span\u003e\u003cspan address=\"10.1016/j.worlddev.2008.01.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIves JD, Messerli B. The Himalayan dilemma: Reconciling development and conservation. Routledge; 2003.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJobe RT, White PS. A new cost-distance model for human accessibility and an evaluation of accessibility bias in permanent vegetation plots in Great Smoky Mountains National Park, USA. J Veg Sci. 2009;20(6):1099\u0026ndash;109.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKellerman A. Distance: A 2024 re-examination. GeoJournal. 2024;89(4):183. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10708-024-11176-6\u003c/span\u003e\u003cspan address=\"10.1007/s10708-024-11176-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKrishi Vigyan Kendra West Kameng. (2024). District profile. District profile. Krishi Vigyan Kendra, West Kameng \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://kvkwestkameng.gov.in/districtprofile.html\u003c/span\u003e\u003cspan address=\"https://kvkwestkameng.gov.in/districtprofile.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (Accessed 12 January 2026).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu L, Li X. GIS and built environment. GeoJournal. 2014;79(2):127\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10708-013-9507-9\u003c/span\u003e\u003cspan address=\"10.1007/s10708-013-9507-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLloyd CB, Mensch BS, Clark WH. The effects of primary school quality on school dropout among Kenyan girls and boys. Comp Educ Rev. 2000;44(2):113\u0026ndash;47. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1086/447600\u003c/span\u003e\u003cspan address=\"10.1086/447600\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMacharia PM, Mumo E, Okiro EA. Modelling geographical accessibility to urban centres in Kenya in 2019. PLoS ONE. 2021;16(5):e0251624. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0251624\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0251624\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeena DK, Tripathi R, Agrawal S. An evaluation of primary schools and their accessibility using GIS techniques: A case study of Prayagraj district. India Geoj. 2023;88(2):1921\u0026ndash;51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10708-022-10715-3\u003c/span\u003e\u003cspan address=\"10.1007/s10708-022-10715-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMinistry of Education, Government of India. (2022). Unified District Information System for Education Plus (UDISE+) 2021\u0026ndash;22: All India report card on school education. Government of India. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.education.gov.in/sites/upload_files/mhrd/files/statistics-new/udise_21_22.pdf\u003c/span\u003e\u003cspan address=\"https://www.education.gov.in/sites/upload_files/mhrd/files/statistics-new/udise_21_22.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (Accessed 11 September 2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeutens T. Accessibility, equity and health care: Review and research directions for transport geographers. J Transp Geogr. 2015;43:14\u0026ndash;27. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jtrangeo.2014.12.006\u003c/span\u003e\u003cspan address=\"10.1016/j.jtrangeo.2014.12.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eO'Sullivan D, Morrison A, Shearer J. Using desktop GIS for the investigation of accessibility by public transport: An isochrone approach. Int J Geogr Inf Sci. 2000;14(1):85\u0026ndash;104. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/136588100240976\u003c/span\u003e\u003cspan address=\"10.1080/136588100240976\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePallathadka A, Pallathadka L, Rao S, Chang H, Van Dommelen D. Using GIS-based spatial analysis to determine urban greenspace accessibility for different racial groups in the backdrop of COVID-19: A case study of four US cities. GeoJournal. 2022;87(6):4879\u0026ndash;99. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10708-021-10538-8\u003c/span\u003e\u003cspan address=\"10.1007/s10708-021-10538-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRay N, Ebener S. AccessMod 3.0: Computing geographic coverage and accessibility to health care services using anisotropic movement of patients. Int J Health Geogr. 2008;7(1):63. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/1476-072X-7-63\u003c/span\u003e\u003cspan address=\"10.1186/1476-072X-7-63\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSiegel S. Nonparametric statistics for the behavioral sciences. McGraw-Hill; 1956.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoukhov A, Pereira RHM, Higgins CD, P\u0026aacute;ez A. A family of accessibility measures derived from spatial interaction principles. PLoS ONE. 2025;20(11):e0335951. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1371/journal.pone.0335951\u003c/span\u003e\u003cspan address=\"10.1371/journal.pone.0335951\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTiwari PC, Tiwari A, Joshi B. Urban growth in Himalaya: Understanding the process and options for sustainable development. J Urban Reg Stud Contemp India. 2018;4(2):15\u0026ndash;26. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.15027/45582\u003c/span\u003e\u003cspan address=\"10.15027/45582\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTobler W. (1993). Three presentations on geographical analysis and modeling: Non-isotropic geographic modeling; speculations on the geometry of geography; and global spatial analysis (NCGIA Technical Report 93\u0026thinsp;\u0026ndash;\u0026thinsp;1). National Center for Geographic Information and Analysis, University of California, Santa Barbara. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://escholarship.org/uc/item/05r820mz\u003c/span\u003e\u003cspan address=\"https://escholarship.org/uc/item/05r820mz\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeiss, D. J., Nelson, A., Gibson, H. S., Temperley, W., Peedell, S., Lieber, A., \u0026hellip;Gething, P. W. (2018). A global map of travel time to cities to assess inequalities in accessibility. Nature, 553(7688), 333\u0026ndash;336. https://doi.org/10.1038/nature25181.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZanaga D, Van De Kerchove R, Daems D, De Keersmaecker W, Brockmann C, Kirches G, Wevers J, Cartus O, Santoro M, Fritz S, Lesiv M, Herold M, Tsendbazar NE, Xu P, Ramoino F, Arino O. (2022). ESA WorldCover 10 m 2021 v200 [Data set]. Zenodo. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5281/zenodo.7254221\u003c/span\u003e\u003cspan address=\"10.5281/zenodo.7254221\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZanaga, D., Van De Kerchove, R., De Keersmaecker, W., Souverijns, N., Brockmann, C.,Quast, R., \u0026hellip; Arino, O. (2021). ESA WorldCover 10 m 2020 v100 (pp. 1\u0026ndash;27). https://doi.org/10.5281/zenodo.5571936.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu Y, Zinda JA, Liu Q, Wang Y, Fu B, Li M. Accessibility of primary schools in rural areas and the impact of topography: A case study in Nanjiang County, China. Land. 2023;12(6):1134. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/land12061134\u003c/span\u003e\u003cspan address=\"10.3390/land12061134\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"discover-global-society","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Discover Global Society](https://www.springer.com/journal/44282)","snPcode":"44282","submissionUrl":"https://submission.nature.com/new-submission/44282/3","title":"Discover Global Society","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Terrain-constrained accessibility, School Accessibility Index (SAI), School dropout, Spatial inequality, Mountain regions","lastPublishedDoi":"10.21203/rs.3.rs-9210282/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9210282/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eTerrain is widely acknowledged as a barrier to schooling in mountainous regions. However, it is rarely modelled as an independent determinant of school accessibility and retention. This study examines the spatial relationship between terrain-induced accessibility and school-level dropout patterns across primary, middle, and secondary and above stages in West Kameng district, Arunachal Pradesh. A GIS-based school accessibility index (SAI) was constructed by integrating slope and land-use resistance within a cost-distance framework. The final analysis includes 41 primary, 58 middle, and 24 secondary and above schools. Accessibility remains valley concentrated. Most primary schools fall within moderate accessibility zones. None are located in very low-accessibility zones. However, dropout at the primary stage does not follow a consistent accessibility gradient. At the middle stage, the dropout rate increases as accessibility decreases. It increases from 3.24% in moderate zones to 9.31% in very low-accessibility zones. However, at the secondary level, the dropout rate remains the lowest in high-accessibility zones. Moderate and low-accessibility schools recorded relatively high rates of dropout. However, the pattern in this stage is less uniform. The results of the Kruskal-Wallis tests do not reveal statistically significant differences across SAI classes. However, descriptive gradients suggest that terrain-related accessibility constraints become more visible with educational progression. The findings indicate that terrain may function as a spatial conditioning factor rather than a direct determinant of dropout. Mobility support and residential planning are therefore critical for sustaining post-primary education in high-relief districts.\u003c/p\u003e","manuscriptTitle":"Analysing the relationship between terrain constrained school accessibility and dropout patterns in West Kameng District Arunachal Pradesh India","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-02 18:41:01","doi":"10.21203/rs.3.rs-9210282/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-04-17T13:17:39+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-08T19:12:57+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-06T14:59:40+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-05T15:36:38+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"105666408892340748889662255216754167152","date":"2026-04-04T04:44:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"66334959913227831904421003865345294565","date":"2026-04-01T05:06:17+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"196778882777260805321543736733813499666","date":"2026-03-31T04:54:51+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"44973639249253863806536889753720295939","date":"2026-03-30T06:42:16+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"76249214269564872962650872860191636351","date":"2026-03-30T05:17:36+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-30T04:41:38+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-28T01:54:26+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-27T10:26:51+00:00","index":"","fulltext":""},{"type":"submitted","content":"Discover Global Society","date":"2026-03-27T10:20:51+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"discover-global-society","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Discover Global Society](https://www.springer.com/journal/44282)","snPcode":"44282","submissionUrl":"https://submission.nature.com/new-submission/44282/3","title":"Discover Global Society","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"747754f5-d478-446e-88f5-a5a2c642ea52","owner":[],"postedDate":"April 2nd, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2026-04-17T13:24:30+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-02 18:41:01","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9210282","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9210282","identity":"rs-9210282","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}