Emergence of spatial heterogeneity in crop water stress under soil drying: Evidence from UAV thermal imagery and physiological measurements

Emergence of spatial heterogeneity in crop water stress under soil drying: Evidence from UAV thermal imagery and physiological measurements | 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 Emergence of spatial heterogeneity in crop water stress under soil drying: Evidence from UAV thermal imagery and physiological measurements Atsushi Sakaguchi, Cheng-Yuan Xu, Haruyuki Fujimaki, Kadambot Siddique This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9449509/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 6 You are reading this latest preprint version Abstract Spatial heterogeneity in crop water stress is a critical but still poorly understood phenomenon at the field scale. This study investigates how such heterogeneity emerges under soil drying conditions using unmanned aerial vehicle (UAV)-based thermal imaging and ground-based physiological measurements. UAV observations over a maize field showed that canopy temperature was relatively uniform under well-watered conditions but became increasingly heterogeneous as soil drying progressed. The relative standard deviation of canopy temperature increased markedly beyond a threshold soil matric potential of approximately − 50 kPa. This trend was not observed under well-irrigated conditions with drip irrigation, indicating that spatial heterogeneity does not inherently exist but rather emerges under water-limited conditions. Ground-based measurements in a sesame field revealed that, as soil drying progressed, the spatial variability of stomatal conductance increased, while its mean value decreased. The increase in variability was reversible following irrigation and was primarily observed at the interplant scale. The observed behavior can be interpreted using the Feddes root water uptake model, in which differences in soil water status do not affect plant water stress under wet conditions but do under dry conditions. Importantly, previous studies have shown that spatial variability in soil water remains substantial under both wet and dry conditions, suggesting that the observed increase in plant-level heterogeneity arises from a nonlinear amplification of pre-existing soil water variability. Crop water stress Spatial heterogeneity Soil matric potential UAV thermal imaging Stomatal conductance Leaf temperature Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 1. Introduction In recent years, increasing efforts have been made to assess crop water stress at the field scale using unmanned aerial vehicles (UAVs) (Santesteban et al., 2017; Park et al., 2021; Liu et al., 2024; Kapari et al., 2025; Payares et al., 2025). Concurrently, the application of UAVs for field water management has advanced rapidly, with Ortega-Farias et al. (2025) summarizing recent research trends. When such water stress information is used for irrigation management, irrigation cannot be applied at the individual plant level; therefore, field-averaged values are typically used as irrigation indicators. However, crop water stress is not spatially uniform within a field. What governs this spatial heterogeneity in crop water stress remains an open question. A plausible explanation is that the underlying drivers can be identified by examining environmental factors that co-vary with changes in heterogeneity. One potential factor is soil moisture within the root zone. In all of the studies cited above that assessed crop water stress, spatial variability in canopy temperature derived from UAV imagery was consistently observed. Although these studies also included soil moisture measurements, the relationship between canopy temperature variability and soil moisture has yet to be clearly established. Moreover, the crop water stress index (CWSI) (Jackson, 1981)—a widely used index—is typically calculated using mean canopy temperature, potentially overlooking information contained in spatial variability. The CWSI was originally developed for individual leaves under controlled conditions and was subsequently extended to the canopy scale using canopy temperature measurements obtained from infrared thermometers viewing the crop canopy, where the mean canopy temperature is used as a representative value (Payero and Irmak, 2006). This approach has since been widely adopted and continues to be used in recent studies (e.g., Romero-Trigueros et al., 2019), implicitly assuming that the crop canopy can be treated as a single, homogeneous “big leaf.” However, it remains unclear whether this assumption holds under field conditions, where spatial variability in canopy temperature is present. Even when the mean canopy temperature is identical, different underlying temperature distributions may correspond to different levels of crop water stress. Clarifying the relationship between spatial variability and crop water stress may provide important insights into the validity of treating the crop canopy as a “big leaf” and, in the context of irrigation management, whether the entire field canopy can be represented as a single functional unit. Therefore, the objectives of this study are to (1) quantify the spatial heterogeneity of crop water stress using UAV-based thermal imagery, (2) examine how this heterogeneity changes with soil drying, and (3) elucidate the physiological mechanisms underlying the observed patterns in canopy temperature and stomatal conductance under contrasting soil moisture conditions. 2. Methodology 2.1 Study sites and experimental design Two field experiments were conducted under dry-season conditions in northern Australia. Experiment 1 was conducted in a maize field at the Frank Wise Institute of Tropical Agriculture, Western Australia (15º06’ S, 128º07’ E). The field area was 14,000 m 2 . Following rainfall events on 1–2 July 2022, no further rainfall occurred during the observation period. From 5 July, the field was divided into a non-irrigated plot and a drip-irrigated plot. The irrigated plot was maintained under non-stress conditions, whereas the non-irrigated plot was allowed to dry progressively. Leaf stomatal conductance and UAV-based canopy temperature were measured in both plots from 5 July to 11 August 2022 as indicators of crop water stress, together with soil water measurements. Experiment 2 was conducted in a sesame field at the Katherine Research Station, Northern Territory, Australia (14º28’ S, 132º16’ E). The field area was 10,000 m 2 . Leaf temperature and stomatal conductance were measured from 8 to 19 September 2023 as indicators of crop water stress, together with soil water measurements. Sprinkler irrigation was applied at 3–4 day intervals, and no rainfall occurred during the observation period. 2.2 Soil water measurements In both experiments, soil matric potential was measured using TEROS 21 sensors (Meter, Pullman, USA) installed at depths of 0.1, 0.2, and 0.3 m at three locations within the field. Measurements were recorded at 15-minute intervals using a ZL6 data logger (Meter, Pullman, USA). Soil matric potential was used as an indicator of root-zone soil water availability. 2.3 UAV-based thermal imaging (Experiment 1) Thermal images of the maize field were acquired using a UAV (Mavic 2 Enterprise Dual, DJI, Shenzhen, China) equipped with a thermal camera (640 × 480 pixels). The UAV was positioned at a height of 42 m above ground level at the center of the non-irrigated and drip-irrigated plots. Images were acquired every 10 minutes at 10:00–10:50, 12:00–12:50, and 14:00–14:50 each day under clear-sky conditions. Images of the two plots were acquired sequentially within a few seconds by moving the UAV between plots and were therefore treated as quasi-simultaneous observations. Prior to analysis, image preprocessing was performed to minimize the influence of non-canopy elements and ensure image consistency. Edge regions corresponding to 15% of each image were removed to avoid geometric distortion near the boundaries. A circular mask centered on the image was then applied to reduce radial distortion. To reduce the influence of soil background and shaded pixels, the lowest 10% of pixel values within each image, corresponding to lower temperatures, were excluded. In the analyzed images, the ground surface was located beneath maize canopies and remained shaded from incoming radiation even at midday. Consequently, these shaded soil pixels corresponded to the lowest temperatures in the images. This thresholding procedure was applied consistently across all images. 2.4 Relative water stress indices (Experiment 1) To evaluate the relationships among soil drying, crop water stress, and spatial variability, relative indices were calculated as ratios of values in the non-irrigated plot to those in the drip-irrigated plot. Irrigation in the drip-irrigated plot was managed to maintain soil matric potential at 0.1 m depth between − 9.8 and − 15.5 kPa (pF 2.0–2.2) using drip tubes with 0.1 m emitter spacing along each maize row. The irrigation amount was adjusted based on soil matric potential measured at 9:00. Because it cannot be assumed that maize in the drip-irrigated plot was completely free from water stress, this plot was used as a reference representing a low or relatively low water-stress condition. Relative stomatal conductance was defined as the ratio of stomatal conductance in the non-irrigated plot to that in the drip-irrigated plot. Stomatal conductance was measured using an SC-1 leaf porometer (Meter, Pullman, USA). Measurements were conducted hourly from 9:00 to 15:59 on five randomly selected leaves in each plot. These leaves were selected from different plants and were the third fully expanded leaf of each plant, and plants in the outermost four rows were not included. Because stomatal conductance exhibits diurnal variation, hourly ratios between plots were calculated. Normalizing stomatal conductance by values under potential transpiration conditions (e.g., in a well-watered reference plot) is commonly used as an indicator of crop water stress. The relative standard deviation (RSD) of canopy temperature was defined as the ratio of the RSD in the non-irrigated plot to that in the drip-irrigated plot. Because RSD does not become zero under non-stressed conditions and varies diurnally, ratios were calculated using paired images acquired at the same time. 2.5 Stomatal conductance measurements (Experiment 2) Stomatal conductance of sesame leaves was measured using an SC-1 leaf porometer (Meter Group, Pullman, USA). Measurements were conducted hourly from 9:00 to 15:59 on 10 randomly selected leaves per observation time. These leaves were selected from different plants and were the third fully expanded leaf of each plant. Different plants were selected at each observation time. Although stomatal conductance is influenced by water stress and other environmental factors, meteorological conditions during the measurement period showed minimal day-to-day variation. Therefore, the hourly mean stomatal conductance of the 10 leaves was assumed to primarily reflect plant responses to soil drying. 2.6 Leaf temperature measurements (Experiment 2) Leaf temperature was measured using a thermal camera (FLIR ONE Pro, Teledyne FLIR, Wilsonville, USA). Thermal images were acquired hourly from 9:00 to 15:59 for 20 randomly selected plants, with images taken from above the canopy. Leaf temperatures were extracted using analysis software (FLIR Tools, Teledyne FLIR, Wilsonville, USA). For each plant, the temperatures of the uppermost 10 fully expanded leaves were recorded. 2.7 Evaluation of spatial heterogeneity Spatial variability of canopy temperature, stomatal conductance, and leaf temperature was quantified using the relative standard deviation (RSD), defined as: $$\:RSD=\frac{\sigma\:}{\mu\:}$$ where σ represents the standard deviation, and µ represents the mean. Although variability can be expressed using standard deviation, RSD was used to enable comparisons among variables with different magnitudes. For canopy temperature, RSD was calculated from the remaining pixel values after applying the image processing steps described in Section 2.3 . For stomatal conductance, RSD was calculated across all sampled leaves at each observation time. For leaf temperature, three types of variability were evaluated: (i) RSD of all leaves in the field, (ii) RSD of mean leaf temperature among plants, obtained by averaging leaf temperature within each plant and then calculating the RSD across all plants, and (iii) within-plant RSD of leaf temperature, obtained by calculating the RSD within each plant and then averaging these values across all plants. Strictly speaking, leaf temperature should be weighted by leaf area when calculating RSD; however, all leaves were assumed to have equal area, and no weighting was applied. These analyses were used to identify the spatial scale at which heterogeneity occurred. 3. Results 3.1 UAV-based observations in the maize field 3.1.1 Field-scale variability observed by UAV thermal imagery Clear differences in the spatial distribution of maize canopy temperature were observed under contrasting soil moisture conditions (Fig. 1 ). Under wet conditions (–15 kPa, pF 2.18), canopy temperature was relatively low and spatially uniform (RSD = 11.3%). In contrast, under dry conditions (–1413 kPa, pF 4.15), canopy temperature increased and exhibited greater spatial heterogeneity (RSD = 19.0%), indicating enhanced variability under soil drying. The soil at 0.1 m depth was consistently very dry, and maize roots were rarely observed at this depth compared with deeper layers. Given the root distribution of maize in soil profile, soil matric potential at 0.3 m depth was used in Figs. 1 – 4 . The RSD of canopy temperature increased with soil drying (Fig. 2 ). This trend was consistently observed across different times of day, with RSD tending to be higher at midday than in the morning, regardless of soil moisture status. 3.1.2 Comparison of water stress indicators with the drip-irrigated plot Figure 2 may reflect factors independent of soil drying that contribute to the increase in RSD. To account for this, Fig. 3 presents the RSD normalized by that of the drip-irrigated plot, using UAV images acquired nearly simultaneously (within a few seconds). A ratio of 1.0 indicates identical RSD values between the two plots. Although a reference line at 1.0 is shown, values may fall below 1.0 when RSD values are similar across plots. The relative RSD of canopy temperature showed little dependence on soil moisture under wet conditions but increased markedly with soil drying, particularly beyond − 50 kPa (pF 2.7). Because root-zone soil matric potential is the primary difference between the two plots, this trend can be attributed to soil drying, indicating that spatial heterogeneity becomes pronounced only beyond a threshold soil moisture condition. Stomatal conductance measured concurrently in both plots is shown in Fig. 4 as a ratio relative to the drip-irrigated plot. Values were compared on hourly, with each value representing the mean of five leaves, and are presented as averages over three time periods. A ratio of 1.0 indicates identical stomatal conductance between plots. Although a reference line at 1.0 is included, values may exceed 1.0 when differences between plots are minimal. Relative stomatal conductance decreased with soil drying, particularly beyond − 50 kPa (pF 2.7), indicating increasing leaf-level water stress in the non-irrigated plot. Both ratios exhibited scatter in their relationships with soil matric potential because the reference values were not strictly constant. However, consistent trends are evident in the upper values in Fig. 3 and the lower values in Fig. 4 . Temporal variation in reference values is likely attributable to fluctuations in transpiration rate and/or the fact that even under drip irrigation, conditions are not entirely free from water stress. Nevertheless, most values in Fig. 3 exceed 1.0, whereas most values in Fig. 4 are below 1.0, indicating that canopy temperature variability is generally greater, and water stress more pronounced, in the non-irrigated plot than in the drip-irrigated plot. Canopy coverage remained approximately constant (71%) during the observation period, as confirmed from visible images acquired simultaneously with the thermal images, indicating that changes in canopy cover did not drive the observed changes in RSD. 3.2 Ground-based observations in the sesame field 3.2.1 Stomatal conductance response to soil drying Because a hardpan layer was present in the shallow soil of the sesame field, and the soil at 0.3 m depth was much harder than that above 0.2 m, soil matric potential at 0.2 m depth was used in Figs. 5 and 6 . Hourly mean stomatal conductance of sesame declined with increasing soil drying (Fig. 5 ). In contrast, the spatial variability of stomatal conductance increased with soil drying (Fig. 6 ). These results indicate a clear inverse relationship between mean stomatal conductance and its spatial variability. Although soil matric potential fluctuated due to repeated irrigation, the relationship between spatial variability and soil matric potential remained consistent. Furthermore, the RSD of stomatal conductance increased with soil drying but decreased following irrigation (Fig. 7 ), demonstrating that the increase in spatial heterogeneity is reversible and closely linked to soil water dynamics. 3.2.2 Scale-dependent variability in leaf temperature The RSD of leaf temperature across all leaves increased with soil drying (Fig. 8 ), consistent with the pattern observed for stomatal conductance (Fig. 6 ). A decrease in stomatal conductance is associated with an increase in leaf temperature due to reduced transpiration. To clarify how leaf temperature variability arises, two components were compared: (i) the RSD of mean leaf temperature among plants and (ii) the within-plant RSD of leaf temperature. As shown in Fig. 8 , the RSD of mean leaf temperature among plants exhibited a trend nearly identical to that of the RSD calculated across all leaves in response to soil drying. In contrast, the within-plant RSD remained nearly constant across soil moisture conditions, indicating no clear relationship with soil drying. These results demonstrate that spatial heterogeneity in canopy temperature arises primarily from differences among plants rather than variability within individual plants. 4. Discussion 4.1 Emergence of spatial heterogeneity under soil drying This study demonstrated that spatial heterogeneity in crop water stress increases as soil drying progresses. The UAV observations showed that canopy temperature was relatively uniform under wet conditions but became increasingly heterogeneous under dry conditions. This trend was quantitatively supported by the increase in RSD with soil drying (Fig. 2 ). More importantly, the ratio of RSD in the non-irrigated plot to that in the drip-irrigated plot also increased with soil drying (Fig. 3 ). Since the drip-irrigated plot was subject to minimal or relatively low water stress, this comparative result provides strong evidence that soil drying was the primary driver of the RSD increase. Together, these results indicate that spatial heterogeneity in crop water stress is not inherent but is amplified as soil water availability declines. As shown in Fig. 2 , RSD tended to be higher at midday than in the morning, regardless of soil moisture conditions, due to differences in plant water status and atmospheric demand. In the morning, plants can maintain relatively high transpiration rates even under dry soil conditions because water is stored in plant tissues from nighttime root water uptake (Caird et al., 2007; Fang et al., 2018). In contrast, higher midday atmospheric evaporative demand intensifies water stress, leading to greater spatial variability among plants. A similar tendency was reported by Sakaguchi et al. (2021), who investigated the relationship between soil matric potential and the ratio of stomatal conductance between water-stressed and non-water-stressed plots in a soybean field at 10:00, 11:00, 13:00 and 14:00. They found that the correlation between stomatal conductance ratio and soil matric potential was weak at 10:00, indicating that soil drying is expressed less strongly as plant water stress in the morning. 4.2 Physiological basis and interpretation based on the Feddes model Ground-based measurements revealed that the observed increase in spatial heterogeneity is closely linked to plant physiological responses. Stomatal conductance decreased with soil drying, particularly beyond − 50 kPa (pF 2.7) (Figs. 4 and 5 ), while its spatial variability increased (Fig. 6 ), indicating that plants respond non-uniformly to soil drying. For maize and sesame, the depths at which soil matric potential was measured differ between figures; however, − 50 kPa (pF 2.7) corresponds to a critical threshold below which plant water uptake becomes increasingly limited. The observed behavior can be interpreted using the Feddes root water uptake model (Fig. 9 ) (Feddes et al., 1978), which describes the relationship between soil matric potential and root water uptake, with uptake becoming limited once soil water potential decreases beyond a threshold (h 3 ) representing the onset of water stress. In the wet range, differences in root-zone soil water status among locations do not lead to differences in plant water stress because root water uptake is not limiting. In contrast, in the dry range, differences in soil water status directly translate into differences in plant water stress. Thus, spatial variability in soil water availability is effectively “masked” under wet conditions but becomes expressed as variability in plant water stress once soil water becomes limiting. Because plant water stress is reflected in stomatal conductance and canopy temperature, spatial variability in these variables increases with soil drying, thereby increasing RSD. Notably, soil water distribution within a field is spatially heterogeneous under both wet and dry conditions. Observations by Kameyama et al. (2019), who installed sensors on a 10 m grid within a 60 m × 60 m soybean field with a light clay texture and Andosol soil type, showed substantial variability in soil matric potential under both conditions. From Fig. 4 , we estimated that when the field-average matric potential (0–0.2 m depth) was approximately − 20 kPa (pF 2.3), individual locations ranged from − 8 kPa (pF 1.9) to − 126 kPa (pF 3.1) (RSD ~ 9.9%). When the field-average value was approximately − 100 kPa (pF 3.0), individual locations ranged from − 63 kPa (pF 2.8) to − 631 kPa (pF 3.8) (RSD ~ 7.6%). These findings indicate that spatial variability in soil water does not necessarily increase with soil drying. Instead, the present results suggest that pre-existing soil water heterogeneity is amplified nonlinearly at the plant level under water-limited conditions. Variability among individual plants may also contribute to the emergence of spatial heterogeneity. First, spatial differences in leaf area index (LAI) among plants can lead to differences in potential transpiration rates. Second, differences in root biomass or rooting depth may result in variation in water uptake capacity. Both mechanisms can generate plant-to-plant differences in the threshold at which transpiration exceeds root water uptake, corresponding to variability in the threshold soil water condition (h 3 ) in the Feddes model (Fig. 9 ). Even under spatially uniform soil matric potential, variability in h 3 among plants may lead to differences in the onset of water stress during the transition from wet to dry conditions, thereby increasing heterogeneity among plants. As soil drying progresses, these differences may result in asynchronous onset of water stress across the field. 4.3 Reversibility and scale-dependent characteristics Although soil matric potential fluctuated due to repeated irrigation, RSD consistently tracked changes in soil water status (Fig. 6 ), indicating that soil water availability, rather than time per se , governs the observed variability. Temporal analysis further showed that the increase in spatial heterogeneity was reversible (Fig. 7 ): variability increased as the soil dried, whereas soil wetting following irrigation reduced variability. This reversibility indicates that the observed heterogeneity reflects dynamic plant responses rather than irreversible physiological damage. Analysis across spatial scales further showed that variability arises primarily at the interplant scale (Fig. 8 ), while within-plant variability remains small, suggesting that spatial heterogeneity originates from differences among plants rather than within individual plants. Under well-watered conditions, even interplant variability remained low, reinforcing that heterogeneity emerges primarily under water-limited conditions. 4.4 Implications for field monitoring and irrigation The observed phenomenon has important implications for field-scale monitoring and irrigation management. For field monitoring, when using UAV thermal imagery or ground-based measurements across multiple plants, spatial heterogeneity should be explicitly considered, particularly under drying conditions. More fundamentally, these findings also have implications for the common “big leaf” assumption, which treats the crop canopy as a single, uniform unit; the observed spatial heterogeneity suggests that this assumption may not hold under drying conditions, where plant-to-plant differences in water stress become significant. In irrigation management, this implies that field-average indicators may overlook localized water stress, leading irrigation thresholds based solely on field-average values to underestimate stress in subsets of plants. Applying irrigation before severe stress develops may therefore improve overall field performance. If spatial heterogeneity is driven by variability in LAI or root traits, reducing such variability through crop management or breeding may enhance field-scale productivity. However, further research is needed to identify the dominant mechanisms underlying the observed heterogeneity. A promising approach is to integrate measurements of sap flow, soil matric potential near individual plants, stomatal conductance, LAI, and root distribution within the same field. Such integrated datasets would enable disentangling the extent to which heterogeneity arises from differences in water supply, plant demand, or root water uptake capacity. 5. Conclusions This study demonstrated that spatial heterogeneity in crop water stress emerges and intensifies as soil dries. The phenomenon was consistently observed across crops and measurement methods, including UAV-derived canopy temperature, ground-based stomatal conductance, and leaf temperature. Heterogeneity primarily occurred at the interplant scale, was reversible following re-watering, and was more pronounced at midday than in the morning. The Feddes root water uptake model provides a useful conceptual framework for interpreting these patterns, highlighting how pre-existing variability in soil water and plant traits can be nonlinearly amplified under water-limited conditions. Declarations Acknowledgments This study was made possible by permissions granted from the Frank Wise Institute of Tropical Agriculture (Western Australia Government) and the Katherine Research Station (Northern Territory Government) to conduct observations at their trial sites. This work was supported by JSPS KAKENHI (Grant No. 21K14941) and partly funded by the Joint Research Program of the Arid Land Research Center, Tottori University (No. 05B2013). Cheng-Yuan Xu’s position is funded by Regional Research Collaboration Program project ‘Research Institute for Northern Agriculture and Drought Resilience’, which is supported by the Australian Department of Education. Funding This work was supported by JSPS KAKENHI (Grant No. 21K14941) and partly funded by the Joint Research Program of the Arid Land Research Center, Tottori University (No. 05B2013). Competing Interests The authors declare no competing interests. Author Contributions A.S. conceived the study, designed the experiments, conducted UAV and field measurements, and performed data analysis. S.X. facilitated field experiments, contributed to experimental design, and provided access to the sesame field in Northern Territory, Australia; he also contributed to the interpretation of the results and provided critical revisions of the manuscript. K.H.M.S. facilitated field experiments, contributed to experimental design, and provided access to the maize field in Western Australia, Australia; he also provided critical revisions of the manuscript. H.F. contributed to the interpretation of the results and provided critical revisions of the manuscript. A.S. wrote the original draft, and all authors reviewed and approved the final manuscript. References Atencia Payares LK, Gomez-del-Campo M, Tarquis AM, Garcia M (2025) Thermal imaging from UAS for estimating crop water status in a Merlot vineyard in semi-arid conditions. Irrig Sci 43:87–103 Caird MA, Richards JH, Donovan LA (2007) Nighttime stomatal conductance and transpiration in C3 and C4 plants. Plant Physiol 143:4–10 Fang W, Lu N, Zhang Y, Jiao L, Fu B (2018) Responses of nighttime sap flow to atmospheric and soil dryness and its potential roles for shrubs on the Loess Plateau of China. J Plant Ecol 11:717–729 Feddes RA, Kowalik PJ, Zaradny H (1978) Simulation of field water use and crop yield. John Wiley & Sons, New York Jackson RD, Idso SB, Reginato RJ, Pinter PJ (1981) Canopy temperature as a crop water stress indicator. Water Resour Res 17:1133–1138 Kameyama K, Miyamoto T, Iwata Y (2019) Determining representative locations based on soil water distribution measurements using a dielectric soil water sensor. IDRE J 309:113–121 Kapari M, Sibanda M, Magidi J, Mabhaudhi T, Mpandeli S, Nhamo L (2025) Assessment of the maize crop water stress index (CWSI) using drone-acquired data across different phenological stages. Drones 9:192 Liu Q, Qu Z, Hu X, Bai Y, Yang W, Yang Y, Bian J, Zhang D, Shi L (2024) Combining UAV remote sensing data to estimate daily-scale crop water stress index: enhancing diagnostic temporal representativeness. Agric Water Manag 305:109130 Ortega-Farias S, Ramírez-Cuesta JM, Nieto H (2025) Recent advances on water management using UAV-based technologies. Irrig Sci 43:1–3 Park S, Ryu D, Fuentes S, Chung H, O’Connell M, Kim J (2021) Dependence of CWSI-based plant water stress estimation on diurnal acquisition times in a nectarine orchard. Remote Sens 13:2775 Payero J, Irmak S (2006) Variable upper and lower crop water stress index baselines for corn and soybean. Irrig Sci 25:21–32 Romero-Trigueros C, Bayona Gambín JM, Nortes Tortosa PA, Alarcón Cabañero JJ, Nicolás Nicolás E (2019) Determination of crop water stress index by infrared thermometry in grapefruit trees irrigated with saline reclaimed water combined with deficit irrigation. Remote Sens 11:757 Sakaguchi A, Tsuji T, Fujii T, Araki H, Takahashi T (2021) Hourly observation and modeling of relationship between dryness of soil and water stress of soybean measured by stomatal conductance at converted field. J Jpn Soc Soil Phys 149:3–12 Santesteban LG, Di Gennaro SF, Herrero-Langreo A, Miranda C, Royo JB, Matese A (2017) High-resolution UAV-based thermal imaging to estimate the instantaneous and seasonal variability of plant water status within a vineyard. Agric Water Manag 183:49–59 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 29 Apr, 2026 Reviewers agreed at journal 29 Apr, 2026 Reviewers invited by journal 27 Apr, 2026 Editor assigned by journal 18 Apr, 2026 Submission checks completed at journal 18 Apr, 2026 First submitted to journal 17 Apr, 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-9449509","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":631952049,"identity":"af3ac410-0e75-4ab0-babd-3e0d439ccbb3","order_by":0,"name":"Atsushi Sakaguchi","email":"data:image/png;base64,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","orcid":"","institution":"Ibaraki University","correspondingAuthor":true,"prefix":"","firstName":"Atsushi","middleName":"","lastName":"Sakaguchi","suffix":""},{"id":631952053,"identity":"304c9d66-0e36-4b41-824f-c64f70e0370a","order_by":1,"name":"Cheng-Yuan Xu","email":"","orcid":"","institution":"Charles Darwin University","correspondingAuthor":false,"prefix":"","firstName":"Cheng-Yuan","middleName":"","lastName":"Xu","suffix":""},{"id":631952057,"identity":"1a97cb07-ddf5-499f-890f-cff01f774d10","order_by":2,"name":"Haruyuki Fujimaki","email":"","orcid":"","institution":"Tottori University","correspondingAuthor":false,"prefix":"","firstName":"Haruyuki","middleName":"","lastName":"Fujimaki","suffix":""},{"id":631952058,"identity":"367d5503-1c55-4ab9-aecb-2672f297c47f","order_by":3,"name":"Kadambot Siddique","email":"","orcid":"","institution":"The University of Western Australia","correspondingAuthor":false,"prefix":"","firstName":"Kadambot","middleName":"","lastName":"Siddique","suffix":""}],"badges":[],"createdAt":"2026-04-17 13:08:44","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9449509/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9449509/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108839218,"identity":"775cc2c9-2077-41d3-848e-7b1231f33aba","added_by":"auto","created_at":"2026-05-09 00:42:31","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":618999,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial distribution of maize canopy temperature under contrasting soil moisture conditions derived from UAV thermal imagery.\u003c/p\u003e\n\u003cp\u003e(a) Wet condition (–15 kPa, pF 2.18), showing relatively low and spatially uniform canopy temperature (relative standard deviation, RSD = 11.3%, mean = 20.4 °C, Jul 11, 09:13).\u003c/p\u003e\n\u003cp\u003e(b) Dry condition (–1413 kPa, pF 4.15), showing higher canopy temperature and increased spatial heterogeneity (RSD = 19.0%, mean = 30.9 °C, Aug 7, 14:11).\u003c/p\u003e\n\u003cp\u003eSoil matric potential was measured at 0.3 m depth. RSD quantifies the spatial heterogeneity of canopy temperature. Both figures represent the same location.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9449509/v1/c0371cea092d32eba4aaa83e.png"},{"id":108839219,"identity":"e7f9e465-37b7-4bd8-b5bc-83b86e21aa31","added_by":"auto","created_at":"2026-05-09 00:42:31","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":2516857,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between soil matric potential measured at 0.3 m depth and the relative standard deviation (RSD) of maize canopy temperature derived from UAV thermal imagery. Each point represents an individual image. Colors indicate time of day (10:00–10:59, 12:00–12:59, 14:00–14:59). RSD quantifies the spatial heterogeneity of canopy temperature. The curves represent quadratic regression fits of RSD for each time period.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-9449509/v1/48bad927719ec24f648728f0.png"},{"id":108977589,"identity":"4b6bab76-8779-4dab-8020-3fbf1c3360eb","added_by":"auto","created_at":"2026-05-11 11:32:14","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":2420074,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between soil matric potential measured at 0.3 m depth in the non-irrigated plot and the relative RSD of maize canopy temperature derived from UAV thermal imagery. Each point represents an individual image. Colors indicate time of day (10:00–10:59, 12:00–12:59, 14:00–14:59). Horizontal solid line indicates relative RSD = 1.0, and vertical dashed line indicates –50 kPa (pF 2.7). Relative RSD is the ratio of RSD in the non-irrigated plot to that in the drip-irrigated plot.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9449509/v1/0242b0de7bc35076cdafaffe.png"},{"id":108839224,"identity":"1e206f9d-6cb2-4e4d-83aa-165763a71907","added_by":"auto","created_at":"2026-05-09 00:42:31","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2252989,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between soil matric potential measured at 0.3 m depth in the non-irrigated plot and relative stomatal conductance of maize. Each point represents average hourly relative stomatal conductance values for each time period. Colors indicate time of day (09:00–10:59, 11:00–13:59, 14:00–15:59). Horizontal solid line indicates a relative stomatal conductance = 1.0, and vertical dashed line indicates –50 kPa (pF 2.7). Relative stomatal conductance is the ratio of mean stomatal conductance of 5 leaves in the non-irrigated plot to that in the drip-irrigated plot.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-9449509/v1/b9a3f52b424e1baf147892bd.png"},{"id":108977835,"identity":"b65e7221-535b-45fe-9570-8b45bbb528f1","added_by":"auto","created_at":"2026-05-11 11:33:07","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":2165703,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between soil matric potential measured at 0.2 m depth and stomatal conductance of sesame. Each point represents the hourly mean stomatal conductance of 10 leaves measured between 9:00 and 15:59. Vertical dashed line indicates –50 kPa (pF 2.7).\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-9449509/v1/ff3e80f83b043f36633f853e.png"},{"id":108839222,"identity":"daa0d87e-fad0-4025-91cf-6246840f3030","added_by":"auto","created_at":"2026-05-09 00:42:31","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":2079809,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between soil matric potential measured at 0.2 m depth and the relative standard deviation (RSD) of stomatal conductance in sesame. Each point represents the hourly RSD of stomatal conductance, calculated from measurements of 10 leaves at each time between 9:00 and 15:59. Vertical dashed line indicates –50 kPa (pF 2.7). RSD quantifies the spatial heterogeneity of stomatal conductance.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-9449509/v1/c9df74b0f083aba55e455afb.png"},{"id":108977133,"identity":"6c3a1aca-4c5b-4782-82e4-9017c1f74981","added_by":"auto","created_at":"2026-05-11 11:30:32","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":2641562,"visible":true,"origin":"","legend":"\u003cp\u003eTemporal variation in the relative standard deviation (RSD) of stomatal conductance in sesame and soil matric potential. Each point represents the hourly RSD of stomatal conductance, calculated from measurements of 10 leaves ateach time between 9:00 and 15:59. RSD quantifies the spatial heterogeneity of stomatal conductance.\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-9449509/v1/aae63b90bdb21ebb95821cbb.png"},{"id":108839223,"identity":"9a3fba4a-7237-41b0-b236-18974c65509e","added_by":"auto","created_at":"2026-05-09 00:42:31","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":2336465,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between soil matric potential measured at 0.2 m depth and the relative standard deviation (RSD) of leaf temperature in sesame at different spatial scales (all leaves, inter-plant, and intra-plant). Each point represents the hourly RSD of leaf temperature, calculated from measurements at each time between 9:00 and 15:59. RSD quantifies the spatial heterogeneity of leaf temperature. The curves represent quadratic regression fits of RSD for the all-leaves and inter-plant scales.\u003c/p\u003e\n\u003cp\u003eAll leaves: RSD of 200 leaves in the field (20 plants × 10 leaves).\u003c/p\u003e\n\u003cp\u003eInter-plant: RSD of plant-mean leaf temperature, obtained by averaging the temperature of 10 leaves within each plant and then calculating the RSD across 20 plants.\u003c/p\u003e\n\u003cp\u003eIntra-plant: Mean within-plant RSD of leaf temperature, obtained by calculating the RSD of 10 leaves within each plant and then averaging these values across 20 plants.\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-9449509/v1/671077b6502faa755bcebfce.png"},{"id":109081050,"identity":"9e76d40b-562f-47ca-b325-e454aa45c1f8","added_by":"auto","created_at":"2026-05-12 11:55:32","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":1651886,"visible":true,"origin":"","legend":"\u003cp\u003eConceptual illustration of spatial variability in root water uptake based on the Feddes model. In the wet range, differences in soil matric potential among locations do not lead to differences in plant water stress because root water uptake is not limiting. In contrast, in the dry range, differences in soil matric potential directly translate into differences in plant water stress.\u003c/p\u003e","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-9449509/v1/195f089e28a4f32ae91b858d.png"},{"id":109204525,"identity":"4a4a2d77-07cb-4858-8c30-578a8919565f","added_by":"auto","created_at":"2026-05-13 15:00:41","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":18858858,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9449509/v1/31799264-5be5-4707-897e-225c6e1d8ef8.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Emergence of spatial heterogeneity in crop water stress under soil drying: Evidence from UAV thermal imagery and physiological measurements","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eIn recent years, increasing efforts have been made to assess crop water stress at the field scale using unmanned aerial vehicles (UAVs) (Santesteban et al., 2017; Park et al., 2021; Liu et al., 2024; Kapari et al., 2025; Payares et al., 2025). Concurrently, the application of UAVs for field water management has advanced rapidly, with Ortega-Farias et al. (2025) summarizing recent research trends. When such water stress information is used for irrigation management, irrigation cannot be applied at the individual plant level; therefore, field-averaged values are typically used as irrigation indicators. However, crop water stress is not spatially uniform within a field. What governs this spatial heterogeneity in crop water stress remains an open question. A plausible explanation is that the underlying drivers can be identified by examining environmental factors that co-vary with changes in heterogeneity. One potential factor is soil moisture within the root zone. In all of the studies cited above that assessed crop water stress, spatial variability in canopy temperature derived from UAV imagery was consistently observed. Although these studies also included soil moisture measurements, the relationship between canopy temperature variability and soil moisture has yet to be clearly established.\u003c/p\u003e \u003cp\u003eMoreover, the crop water stress index (CWSI) (Jackson, 1981)\u0026mdash;a widely used index\u0026mdash;is typically calculated using mean canopy temperature, potentially overlooking information contained in spatial variability. The CWSI was originally developed for individual leaves under controlled conditions and was subsequently extended to the canopy scale using canopy temperature measurements obtained from infrared thermometers viewing the crop canopy, where the mean canopy temperature is used as a representative value (Payero and Irmak, 2006). This approach has since been widely adopted and continues to be used in recent studies (e.g., Romero-Trigueros et al., 2019), implicitly assuming that the crop canopy can be treated as a single, homogeneous \u0026ldquo;big leaf.\u0026rdquo;\u003c/p\u003e \u003cp\u003eHowever, it remains unclear whether this assumption holds under field conditions, where spatial variability in canopy temperature is present. Even when the mean canopy temperature is identical, different underlying temperature distributions may correspond to different levels of crop water stress. Clarifying the relationship between spatial variability and crop water stress may provide important insights into the validity of treating the crop canopy as a \u0026ldquo;big leaf\u0026rdquo; and, in the context of irrigation management, whether the entire field canopy can be represented as a single functional unit.\u003c/p\u003e \u003cp\u003eTherefore, the objectives of this study are to (1) quantify the spatial heterogeneity of crop water stress using UAV-based thermal imagery, (2) examine how this heterogeneity changes with soil drying, and (3) elucidate the physiological mechanisms underlying the observed patterns in canopy temperature and stomatal conductance under contrasting soil moisture conditions.\u003c/p\u003e"},{"header":"2. Methodology","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study sites and experimental design\u003c/h2\u003e \u003cp\u003eTwo field experiments were conducted under dry-season conditions in northern Australia. Experiment 1 was conducted in a maize field at the Frank Wise Institute of Tropical Agriculture, Western Australia (15\u0026ordm;06\u0026rsquo; S, 128\u0026ordm;07\u0026rsquo; E). The field area was 14,000 m\u003csup\u003e2\u003c/sup\u003e. Following rainfall events on 1\u0026ndash;2 July 2022, no further rainfall occurred during the observation period. From 5 July, the field was divided into a non-irrigated plot and a drip-irrigated plot. The irrigated plot was maintained under non-stress conditions, whereas the non-irrigated plot was allowed to dry progressively. Leaf stomatal conductance and UAV-based canopy temperature were measured in both plots from 5 July to 11 August 2022 as indicators of crop water stress, together with soil water measurements. Experiment 2 was conducted in a sesame field at the Katherine Research Station, Northern Territory, Australia (14\u0026ordm;28\u0026rsquo; S, 132\u0026ordm;16\u0026rsquo; E). The field area was 10,000 m\u003csup\u003e2\u003c/sup\u003e. Leaf temperature and stomatal conductance were measured from 8 to 19 September 2023 as indicators of crop water stress, together with soil water measurements. Sprinkler irrigation was applied at 3\u0026ndash;4 day intervals, and no rainfall occurred during the observation period.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Soil water measurements\u003c/h2\u003e \u003cp\u003eIn both experiments, soil matric potential was measured using TEROS 21 sensors (Meter, Pullman, USA) installed at depths of 0.1, 0.2, and 0.3 m at three locations within the field. Measurements were recorded at 15-minute intervals using a ZL6 data logger (Meter, Pullman, USA). Soil matric potential was used as an indicator of root-zone soil water availability.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 UAV-based thermal imaging (Experiment 1)\u003c/h2\u003e \u003cp\u003eThermal images of the maize field were acquired using a UAV (Mavic 2 Enterprise Dual, DJI, Shenzhen, China) equipped with a thermal camera (640 \u0026times; 480 pixels). The UAV was positioned at a height of 42 m above ground level at the center of the non-irrigated and drip-irrigated plots. Images were acquired every 10 minutes at 10:00\u0026ndash;10:50, 12:00\u0026ndash;12:50, and 14:00\u0026ndash;14:50 each day under clear-sky conditions. Images of the two plots were acquired sequentially within a few seconds by moving the UAV between plots and were therefore treated as quasi-simultaneous observations. Prior to analysis, image preprocessing was performed to minimize the influence of non-canopy elements and ensure image consistency. Edge regions corresponding to 15% of each image were removed to avoid geometric distortion near the boundaries. A circular mask centered on the image was then applied to reduce radial distortion. To reduce the influence of soil background and shaded pixels, the lowest 10% of pixel values within each image, corresponding to lower temperatures, were excluded. In the analyzed images, the ground surface was located beneath maize canopies and remained shaded from incoming radiation even at midday. Consequently, these shaded soil pixels corresponded to the lowest temperatures in the images. This thresholding procedure was applied consistently across all images.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Relative water stress indices (Experiment 1)\u003c/h2\u003e \u003cp\u003eTo evaluate the relationships among soil drying, crop water stress, and spatial variability, relative indices were calculated as ratios of values in the non-irrigated plot to those in the drip-irrigated plot. Irrigation in the drip-irrigated plot was managed to maintain soil matric potential at 0.1 m depth between \u0026minus;\u0026thinsp;9.8 and \u0026minus;\u0026thinsp;15.5 kPa (pF 2.0\u0026ndash;2.2) using drip tubes with 0.1 m emitter spacing along each maize row. The irrigation amount was adjusted based on soil matric potential measured at 9:00. Because it cannot be assumed that maize in the drip-irrigated plot was completely free from water stress, this plot was used as a reference representing a low or relatively low water-stress condition.\u003c/p\u003e \u003cp\u003eRelative stomatal conductance was defined as the ratio of stomatal conductance in the non-irrigated plot to that in the drip-irrigated plot. Stomatal conductance was measured using an SC-1 leaf porometer (Meter, Pullman, USA). Measurements were conducted hourly from 9:00 to 15:59 on five randomly selected leaves in each plot. These leaves were selected from different plants and were the third fully expanded leaf of each plant, and plants in the outermost four rows were not included. Because stomatal conductance exhibits diurnal variation, hourly ratios between plots were calculated. Normalizing stomatal conductance by values under potential transpiration conditions (e.g., in a well-watered reference plot) is commonly used as an indicator of crop water stress.\u003c/p\u003e \u003cp\u003eThe relative standard deviation (RSD) of canopy temperature was defined as the ratio of the RSD in the non-irrigated plot to that in the drip-irrigated plot. Because RSD does not become zero under non-stressed conditions and varies diurnally, ratios were calculated using paired images acquired at the same time.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Stomatal conductance measurements (Experiment 2)\u003c/h2\u003e \u003cp\u003eStomatal conductance of sesame leaves was measured using an SC-1 leaf porometer (Meter Group, Pullman, USA). Measurements were conducted hourly from 9:00 to 15:59 on 10 randomly selected leaves per observation time. These leaves were selected from different plants and were the third fully expanded leaf of each plant. Different plants were selected at each observation time. Although stomatal conductance is influenced by water stress and other environmental factors, meteorological conditions during the measurement period showed minimal day-to-day variation. Therefore, the hourly mean stomatal conductance of the 10 leaves was assumed to primarily reflect plant responses to soil drying.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Leaf temperature measurements (Experiment 2)\u003c/h2\u003e \u003cp\u003eLeaf temperature was measured using a thermal camera (FLIR ONE Pro, Teledyne FLIR, Wilsonville, USA). Thermal images were acquired hourly from 9:00 to 15:59 for 20 randomly selected plants, with images taken from above the canopy. Leaf temperatures were extracted using analysis software (FLIR Tools, Teledyne FLIR, Wilsonville, USA). For each plant, the temperatures of the uppermost 10 fully expanded leaves were recorded.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7 Evaluation of spatial heterogeneity\u003c/h2\u003e \u003cp\u003eSpatial variability of canopy temperature, stomatal conductance, and leaf temperature was quantified using the relative standard deviation (RSD), defined as:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:RSD=\\frac{\\sigma\\:}{\\mu\\:}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere σ represents the standard deviation, and \u0026micro; represents the mean. Although variability can be expressed using standard deviation, RSD was used to enable comparisons among variables with different magnitudes. For canopy temperature, RSD was calculated from the remaining pixel values after applying the image processing steps described in Section \u003cspan refid=\"Sec5\" class=\"InternalRef\"\u003e2.3\u003c/span\u003e. For stomatal conductance, RSD was calculated across all sampled leaves at each observation time. For leaf temperature, three types of variability were evaluated: (i) RSD of all leaves in the field, (ii) RSD of mean leaf temperature among plants, obtained by averaging leaf temperature within each plant and then calculating the RSD across all plants, and (iii) within-plant RSD of leaf temperature, obtained by calculating the RSD within each plant and then averaging these values across all plants. Strictly speaking, leaf temperature should be weighted by leaf area when calculating RSD; however, all leaves were assumed to have equal area, and no weighting was applied. These analyses were used to identify the spatial scale at which heterogeneity occurred.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e3.1 UAV-based observations in the maize field\u003c/h2\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e3.1.1 Field-scale variability observed by UAV thermal imagery\u003c/h2\u003e \u003cp\u003eClear differences in the spatial distribution of maize canopy temperature were observed under contrasting soil moisture conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Under wet conditions (\u0026ndash;15 kPa, pF 2.18), canopy temperature was relatively low and spatially uniform (RSD\u0026thinsp;=\u0026thinsp;11.3%). In contrast, under dry conditions (\u0026ndash;1413 kPa, pF 4.15), canopy temperature increased and exhibited greater spatial heterogeneity (RSD\u0026thinsp;=\u0026thinsp;19.0%), indicating enhanced variability under soil drying. The soil at 0.1 m depth was consistently very dry, and maize roots were rarely observed at this depth compared with deeper layers. Given the root distribution of maize in soil profile, soil matric potential at 0.3 m depth was used in Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e. The RSD of canopy temperature increased with soil drying (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). This trend was consistently observed across different times of day, with RSD tending to be higher at midday than in the morning, regardless of soil moisture status.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e \u003ch2\u003e3.1.2 Comparison of water stress indicators with the drip-irrigated plot\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e may reflect factors independent of soil drying that contribute to the increase in RSD. To account for this, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003e presents the RSD normalized by that of the drip-irrigated plot, using UAV images acquired nearly simultaneously (within a few seconds). A ratio of 1.0 indicates identical RSD values between the two plots. Although a reference line at 1.0 is shown, values may fall below 1.0 when RSD values are similar across plots. The relative RSD of canopy temperature showed little dependence on soil moisture under wet conditions but increased markedly with soil drying, particularly beyond \u0026minus;\u0026thinsp;50 kPa (pF 2.7). Because root-zone soil matric potential is the primary difference between the two plots, this trend can be attributed to soil drying, indicating that spatial heterogeneity becomes pronounced only beyond a threshold soil moisture condition. Stomatal conductance measured concurrently in both plots is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e as a ratio relative to the drip-irrigated plot. Values were compared on hourly, with each value representing the mean of five leaves, and are presented as averages over three time periods. A ratio of 1.0 indicates identical stomatal conductance between plots. Although a reference line at 1.0 is included, values may exceed 1.0 when differences between plots are minimal. Relative stomatal conductance decreased with soil drying, particularly beyond \u0026minus;\u0026thinsp;50 kPa (pF 2.7), indicating increasing leaf-level water stress in the non-irrigated plot. Both ratios exhibited scatter in their relationships with soil matric potential because the reference values were not strictly constant. However, consistent trends are evident in the upper values in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003e and the lower values in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e. Temporal variation in reference values is likely attributable to fluctuations in transpiration rate and/or the fact that even under drip irrigation, conditions are not entirely free from water stress. Nevertheless, most values in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003e exceed 1.0, whereas most values in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e are below 1.0, indicating that canopy temperature variability is generally greater, and water stress more pronounced, in the non-irrigated plot than in the drip-irrigated plot. Canopy coverage remained approximately constant (71%) during the observation period, as confirmed from visible images acquired simultaneously with the thermal images, indicating that changes in canopy cover did not drive the observed changes in RSD.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Ground-based observations in the sesame field\u003c/h2\u003e \u003cdiv id=\"Sec15\" class=\"Section3\"\u003e \u003ch2\u003e3.2.1 Stomatal conductance response to soil drying\u003c/h2\u003e \u003cp\u003eBecause a hardpan layer was present in the shallow soil of the sesame field, and the soil at 0.3 m depth was much harder than that above 0.2 m, soil matric potential at 0.2 m depth was used in Figs.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e and \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e. Hourly mean stomatal conductance of sesame declined with increasing soil drying (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). In contrast, the spatial variability of stomatal conductance increased with soil drying (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). These results indicate a clear inverse relationship between mean stomatal conductance and its spatial variability. Although soil matric potential fluctuated due to repeated irrigation, the relationship between spatial variability and soil matric potential remained consistent. Furthermore, the RSD of stomatal conductance increased with soil drying but decreased following irrigation (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e), demonstrating that the increase in spatial heterogeneity is reversible and closely linked to soil water dynamics.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section3\"\u003e \u003ch2\u003e3.2.2 Scale-dependent variability in leaf temperature\u003c/h2\u003e \u003cp\u003eThe RSD of leaf temperature across all leaves increased with soil drying (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e), consistent with the pattern observed for stomatal conductance (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). A decrease in stomatal conductance is associated with an increase in leaf temperature due to reduced transpiration. To clarify how leaf temperature variability arises, two components were compared: (i) the RSD of mean leaf temperature among plants and (ii) the within-plant RSD of leaf temperature. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e, the RSD of mean leaf temperature among plants exhibited a trend nearly identical to that of the RSD calculated across all leaves in response to soil drying. In contrast, the within-plant RSD remained nearly constant across soil moisture conditions, indicating no clear relationship with soil drying. These results demonstrate that spatial heterogeneity in canopy temperature arises primarily from differences among plants rather than variability within individual plants.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Emergence of spatial heterogeneity under soil drying\u003c/h2\u003e \u003cp\u003eThis study demonstrated that spatial heterogeneity in crop water stress increases as soil drying progresses. The UAV observations showed that canopy temperature was relatively uniform under wet conditions but became increasingly heterogeneous under dry conditions. This trend was quantitatively supported by the increase in RSD with soil drying (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). More importantly, the ratio of RSD in the non-irrigated plot to that in the drip-irrigated plot also increased with soil drying (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Since the drip-irrigated plot was subject to minimal or relatively low water stress, this comparative result provides strong evidence that soil drying was the primary driver of the RSD increase. Together, these results indicate that spatial heterogeneity in crop water stress is not inherent but is amplified as soil water availability declines.\u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, RSD tended to be higher at midday than in the morning, regardless of soil moisture conditions, due to differences in plant water status and atmospheric demand. In the morning, plants can maintain relatively high transpiration rates even under dry soil conditions because water is stored in plant tissues from nighttime root water uptake (Caird et al., 2007; Fang et al., 2018). In contrast, higher midday atmospheric evaporative demand intensifies water stress, leading to greater spatial variability among plants. A similar tendency was reported by Sakaguchi et al. (2021), who investigated the relationship between soil matric potential and the ratio of stomatal conductance between water-stressed and non-water-stressed plots in a soybean field at 10:00, 11:00, 13:00 and 14:00. They found that the correlation between stomatal conductance ratio and soil matric potential was weak at 10:00, indicating that soil drying is expressed less strongly as plant water stress in the morning.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Physiological basis and interpretation based on the Feddes model\u003c/h2\u003e \u003cp\u003eGround-based measurements revealed that the observed increase in spatial heterogeneity is closely linked to plant physiological responses. Stomatal conductance decreased with soil drying, particularly beyond \u0026minus;\u0026thinsp;50 kPa (pF 2.7) (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e), while its spatial variability increased (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e), indicating that plants respond non-uniformly to soil drying. For maize and sesame, the depths at which soil matric potential was measured differ between figures; however, \u0026minus;\u0026thinsp;50 kPa (pF 2.7) corresponds to a critical threshold below which plant water uptake becomes increasingly limited. The observed behavior can be interpreted using the Feddes root water uptake model (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e) (Feddes et al., 1978), which describes the relationship between soil matric potential and root water uptake, with uptake becoming limited once soil water potential decreases beyond a threshold (h\u003csub\u003e3\u003c/sub\u003e) representing the onset of water stress. In the wet range, differences in root-zone soil water status among locations do not lead to differences in plant water stress because root water uptake is not limiting. In contrast, in the dry range, differences in soil water status directly translate into differences in plant water stress. Thus, spatial variability in soil water availability is effectively \u0026ldquo;masked\u0026rdquo; under wet conditions but becomes expressed as variability in plant water stress once soil water becomes limiting. Because plant water stress is reflected in stomatal conductance and canopy temperature, spatial variability in these variables increases with soil drying, thereby increasing RSD.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eNotably, soil water distribution within a field is spatially heterogeneous under both wet and dry conditions. Observations by Kameyama et al. (2019), who installed sensors on a 10 m grid within a 60 m \u0026times; 60 m soybean field with a light clay texture and Andosol soil type, showed substantial variability in soil matric potential under both conditions. From Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e, we estimated that when the field-average matric potential (0\u0026ndash;0.2 m depth) was approximately \u0026minus;\u0026thinsp;20 kPa (pF 2.3), individual locations ranged from \u0026minus;\u0026thinsp;8 kPa (pF 1.9) to \u0026minus;\u0026thinsp;126 kPa (pF 3.1) (RSD\u0026thinsp;~\u0026thinsp;9.9%). When the field-average value was approximately \u0026minus;\u0026thinsp;100 kPa (pF 3.0), individual locations ranged from \u0026minus;\u0026thinsp;63 kPa (pF 2.8) to \u0026minus;\u0026thinsp;631 kPa (pF 3.8) (RSD\u0026thinsp;~\u0026thinsp;7.6%). These findings indicate that spatial variability in soil water does not necessarily increase with soil drying. Instead, the present results suggest that pre-existing soil water heterogeneity is amplified nonlinearly at the plant level under water-limited conditions.\u003c/p\u003e \u003cp\u003eVariability among individual plants may also contribute to the emergence of spatial heterogeneity. First, spatial differences in leaf area index (LAI) among plants can lead to differences in potential transpiration rates. Second, differences in root biomass or rooting depth may result in variation in water uptake capacity. Both mechanisms can generate plant-to-plant differences in the threshold at which transpiration exceeds root water uptake, corresponding to variability in the threshold soil water condition (h\u003csub\u003e3\u003c/sub\u003e) in the Feddes model (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e). Even under spatially uniform soil matric potential, variability in h\u003csub\u003e3\u003c/sub\u003e among plants may lead to differences in the onset of water stress during the transition from wet to dry conditions, thereby increasing heterogeneity among plants. As soil drying progresses, these differences may result in asynchronous onset of water stress across the field.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Reversibility and scale-dependent characteristics\u003c/h2\u003e \u003cp\u003eAlthough soil matric potential fluctuated due to repeated irrigation, RSD consistently tracked changes in soil water status (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e), indicating that soil water availability, rather than time \u003cem\u003eper se\u003c/em\u003e, governs the observed variability. Temporal analysis further showed that the increase in spatial heterogeneity was reversible (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e): variability increased as the soil dried, whereas soil wetting following irrigation reduced variability. This reversibility indicates that the observed heterogeneity reflects dynamic plant responses rather than irreversible physiological damage. Analysis across spatial scales further showed that variability arises primarily at the interplant scale (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e), while within-plant variability remains small, suggesting that spatial heterogeneity originates from differences among plants rather than within individual plants. Under well-watered conditions, even interplant variability remained low, reinforcing that heterogeneity emerges primarily under water-limited conditions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Implications for field monitoring and irrigation\u003c/h2\u003e \u003cp\u003eThe observed phenomenon has important implications for field-scale monitoring and irrigation management. For field monitoring, when using UAV thermal imagery or ground-based measurements across multiple plants, spatial heterogeneity should be explicitly considered, particularly under drying conditions. More fundamentally, these findings also have implications for the common \u0026ldquo;big leaf\u0026rdquo; assumption, which treats the crop canopy as a single, uniform unit; the observed spatial heterogeneity suggests that this assumption may not hold under drying conditions, where plant-to-plant differences in water stress become significant. In irrigation management, this implies that field-average indicators may overlook localized water stress, leading irrigation thresholds based solely on field-average values to underestimate stress in subsets of plants. Applying irrigation before severe stress develops may therefore improve overall field performance. If spatial heterogeneity is driven by variability in LAI or root traits, reducing such variability through crop management or breeding may enhance field-scale productivity. However, further research is needed to identify the dominant mechanisms underlying the observed heterogeneity.\u003c/p\u003e \u003cp\u003eA promising approach is to integrate measurements of sap flow, soil matric potential near individual plants, stomatal conductance, LAI, and root distribution within the same field. Such integrated datasets would enable disentangling the extent to which heterogeneity arises from differences in water supply, plant demand, or root water uptake capacity.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003eThis study demonstrated that spatial heterogeneity in crop water stress emerges and intensifies as soil dries. The phenomenon was consistently observed across crops and measurement methods, including UAV-derived canopy temperature, ground-based stomatal conductance, and leaf temperature. Heterogeneity primarily occurred at the interplant scale, was reversible following re-watering, and was more pronounced at midday than in the morning. The Feddes root water uptake model provides a useful conceptual framework for interpreting these patterns, highlighting how pre-existing variability in soil water and plant traits can be nonlinearly amplified under water-limited conditions.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was made possible by permissions granted from the Frank Wise Institute of Tropical Agriculture (Western Australia Government) and the Katherine Research Station (Northern Territory Government) to conduct observations at their trial sites. This work was supported by JSPS KAKENHI (Grant No. 21K14941) and partly funded by the Joint Research Program of the Arid Land Research Center, Tottori University (No. 05B2013). Cheng-Yuan Xu\u0026rsquo;s position is funded by Regional Research Collaboration Program project \u0026lsquo;Research Institute for Northern Agriculture and Drought Resilience\u0026rsquo;, which is supported by the Australian Department of Education.\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by JSPS KAKENHI (Grant No. 21K14941) and partly funded by the Joint Research Program of the Arid Land Research Center, Tottori University (No. 05B2013).\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA.S. conceived the study, designed the experiments, conducted UAV and field measurements, and performed data analysis.\u003c/p\u003e\n\u003cp\u003eS.X. facilitated field experiments, contributed to experimental design, and provided access to the sesame field in Northern Territory, Australia; he also contributed to the interpretation of the results and provided critical revisions of the manuscript.\u003c/p\u003e\n\u003cp\u003eK.H.M.S. facilitated field experiments, contributed to experimental design, and provided access to the maize field in Western Australia, Australia; he also provided critical revisions of the manuscript.\u003c/p\u003e\n\u003cp\u003eH.F. contributed to the interpretation of the results and provided critical revisions of the manuscript.\u003c/p\u003e\n\u003cp\u003eA.S. wrote the original draft, and all authors reviewed and approved the final manuscript.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAtencia Payares LK, Gomez-del-Campo M, Tarquis AM, Garcia M (2025) Thermal imaging from UAS for estimating crop water status in a Merlot vineyard in semi-arid conditions. Irrig Sci 43:87\u0026ndash;103\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCaird MA, Richards JH, Donovan LA (2007) Nighttime stomatal conductance and transpiration in C3 and C4 plants. Plant Physiol 143:4\u0026ndash;10\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFang W, Lu N, Zhang Y, Jiao L, Fu B (2018) Responses of nighttime sap flow to atmospheric and soil dryness and its potential roles for shrubs on the Loess Plateau of China. J Plant Ecol 11:717\u0026ndash;729\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeddes RA, Kowalik PJ, Zaradny H (1978) Simulation of field water use and crop yield. John Wiley \u0026amp; Sons, New York\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJackson RD, Idso SB, Reginato RJ, Pinter PJ (1981) Canopy temperature as a crop water stress indicator. Water Resour Res 17:1133\u0026ndash;1138\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKameyama K, Miyamoto T, Iwata Y (2019) Determining representative locations based on soil water distribution measurements using a dielectric soil water sensor. IDRE J 309:113\u0026ndash;121\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKapari M, Sibanda M, Magidi J, Mabhaudhi T, Mpandeli S, Nhamo L (2025) Assessment of the maize crop water stress index (CWSI) using drone-acquired data across different phenological stages. Drones 9:192\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Q, Qu Z, Hu X, Bai Y, Yang W, Yang Y, Bian J, Zhang D, Shi L (2024) Combining UAV remote sensing data to estimate daily-scale crop water stress index: enhancing diagnostic temporal representativeness. Agric Water Manag 305:109130\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrtega-Farias S, Ram\u0026iacute;rez-Cuesta JM, Nieto H (2025) Recent advances on water management using UAV-based technologies. Irrig Sci 43:1\u0026ndash;3\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePark S, Ryu D, Fuentes S, Chung H, O\u0026rsquo;Connell M, Kim J (2021) Dependence of CWSI-based plant water stress estimation on diurnal acquisition times in a nectarine orchard. Remote Sens 13:2775\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePayero J, Irmak S (2006) Variable upper and lower crop water stress index baselines for corn and soybean. Irrig Sci 25:21\u0026ndash;32\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRomero-Trigueros C, Bayona Gamb\u0026iacute;n JM, Nortes Tortosa PA, Alarc\u0026oacute;n Caba\u0026ntilde;ero JJ, Nicol\u0026aacute;s Nicol\u0026aacute;s E (2019) Determination of crop water stress index by infrared thermometry in grapefruit trees irrigated with saline reclaimed water combined with deficit irrigation. Remote Sens 11:757\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSakaguchi A, Tsuji T, Fujii T, Araki H, Takahashi T (2021) Hourly observation and modeling of relationship between dryness of soil and water stress of soybean measured by stomatal conductance at converted field. J Jpn Soc Soil Phys 149:3\u0026ndash;12\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSantesteban LG, Di Gennaro SF, Herrero-Langreo A, Miranda C, Royo JB, Matese A (2017) High-resolution UAV-based thermal imaging to estimate the instantaneous and seasonal variability of plant water status within a vineyard. Agric Water Manag 183:49\u0026ndash;59\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":"irrigation-science","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"irsc","sideBox":"Learn more about [Irrigation Science](http://link.springer.com/journal/271)","snPcode":"271","submissionUrl":"https://submission.nature.com/new-submission/271/3","title":"Irrigation Science","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Crop water stress, Spatial heterogeneity, Soil matric potential, UAV thermal imaging, Stomatal conductance, Leaf temperature","lastPublishedDoi":"10.21203/rs.3.rs-9449509/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9449509/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSpatial heterogeneity in crop water stress is a critical but still poorly understood phenomenon at the field scale. This study investigates how such heterogeneity emerges under soil drying conditions using unmanned aerial vehicle (UAV)-based thermal imaging and ground-based physiological measurements.\u003c/p\u003e \u003cp\u003eUAV observations over a maize field showed that canopy temperature was relatively uniform under well-watered conditions but became increasingly heterogeneous as soil drying progressed. The relative standard deviation of canopy temperature increased markedly beyond a threshold soil matric potential of approximately \u0026minus;\u0026thinsp;50 kPa. This trend was not observed under well-irrigated conditions with drip irrigation, indicating that spatial heterogeneity does not inherently exist but rather emerges under water-limited conditions.\u003c/p\u003e \u003cp\u003eGround-based measurements in a sesame field revealed that, as soil drying progressed, the spatial variability of stomatal conductance increased, while its mean value decreased. The increase in variability was reversible following irrigation and was primarily observed at the interplant scale.\u003c/p\u003e \u003cp\u003eThe observed behavior can be interpreted using the Feddes root water uptake model, in which differences in soil water status do not affect plant water stress under wet conditions but do under dry conditions. Importantly, previous studies have shown that spatial variability in soil water remains substantial under both wet and dry conditions, suggesting that the observed increase in plant-level heterogeneity arises from a nonlinear amplification of pre-existing soil water variability.\u003c/p\u003e","manuscriptTitle":"Emergence of spatial heterogeneity in crop water stress under soil drying: Evidence from UAV thermal imagery and physiological measurements","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-09 00:42:15","doi":"10.21203/rs.3.rs-9449509/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"19111475677098192176907938244244859225","date":"2026-04-29T17:47:53+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"88282799808688325746146315519081823064","date":"2026-04-29T09:45:43+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-27T08:46:11+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-18T13:26:59+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-18T13:26:30+00:00","index":"","fulltext":""},{"type":"submitted","content":"Irrigation Science","date":"2026-04-17T13:01:22+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"irrigation-science","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"irsc","sideBox":"Learn more about [Irrigation Science](http://link.springer.com/journal/271)","snPcode":"271","submissionUrl":"https://submission.nature.com/new-submission/271/3","title":"Irrigation Science","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"2524f883-786d-4591-8c0c-be8669be30db","owner":[],"postedDate":"May 9th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"19111475677098192176907938244244859225","date":"2026-04-29T17:47:53+00:00","index":18,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-09T00:42:15+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-09 00:42:15","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9449509","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9449509","identity":"rs-9449509","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}