Infrared Land Surface Emissivity Dynamics in the Taklimakan Desert : Spatiotemporal Patterns and Key Drivers

Infrared Land Surface Emissivity Dynamics in the Taklimakan Desert : Spatiotemporal Patterns and Key Drivers | 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 Article Infrared Land Surface Emissivity Dynamics in the Taklimakan Desert : Spatiotemporal Patterns and Key Drivers Yufen Ma, Kang Zeng, Ailiyaer Aihaiti, Junjian LIU, Zonghui LIU This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7587474/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 15 Jan, 2026 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract This investigation systematically quantifies the spatiotemporal patterns and governing mechanisms of land surface emissivity (LSE) across three infrared wavelengths (8.3, 8.6, and 9.1 µm) in the hyper-arid Taklimakan Desert using 23-year satellite records (2001–2023). Our analysis reveals several key findings: (1) Despite theoretical sensitivity to soil moisture, LSE exhibits a paradoxical decadal increase (0.12 ± 0.03 decade⁻¹) concurrent with regional drying (-0.15 g/kg decade⁻¹), demonstrating thermal processes dominate 68 ± 7% of variability through particle expansion/contraction cycles; (2) Surface temperature exerts independent control, reducing emissivity by 0.0029 ± 0.0012 per 1°C, with maximum sensitivity at 9.1 µm (-0.0035 ± 0.0015); (3) Spectral analysis identifies wavelength-specific responses—the 8.6 µm band displays highest interannual stability (CV = 1.1 ± 0.3%), while 8.3 µm shows greatest surface sensitivity (CV = 2.9 ± 0.5%), with summer peaks (0.89 ± 0.02) amplified by aeolian processes in central dunes (ΔLSE > 0.07). These findings redefine LSE controls in hyper-arid environments through thermal-aeolian coupling mechanisms, providing critical constraints for desertification-climate feedback models. Earth and environmental sciences/Climate sciences Earth and environmental sciences/Environmental sciences Taklimakan Desert Infrared Emissivity thermal dynamics aeolian processes remote sensing validation Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 1. Introduction Land surface emissivity (LSE) plays a pivotal role in governing thermal radiation exchange in hyper-arid regions, yet its spatiotemporal controls remain incompletely understood under extreme dryness. The Taklimakan Desert, characterized by persistent aridity and active aeolian processes, serves as an ideal natural laboratory to examine thermal versus hydrological drivers of LSE beyond diurnal timescales. Recent advances in remote sensing technology have enabled high-resolution analysis of emissivity variations across multiple spatiotemporal scales (Jin and Liang, 2006 ). Previous research indicates LSE sensitivity to surface characteristics including soil moisture, vegetation cover, and mineral composition, as well as atmospheric conditions such as water vapor content and aerosol loading (Li, 2013). In hyper-arid environments like the Taklimakan Desert, where surface heterogeneity is pronounced (Zhang, 2015), quantifying these variations is essential for refining climate models, improving land surface temperature retrievals, and enhancing environmental monitoring capabilities (Chen, 2019). Existing studies have established the significant impact of soil moisture on LSE, particularly in desert ecosystems (Mira, 2020; Hulley, 2020). Laboratory and satellite-based analyses demonstrate that even minor fluctuations in soil moisture can induce detectable emissivity changes (Ma, 2023), with cascading effects on energy partitioning and surface heat fluxes (Ogawa, 2008). While vegetation is sparse in deserts, its presence—such as in oases and riparian zones—modifies LSE through higher emissivity signatures (Sobrino, 2012; French, 2012). Additionally, atmospheric variability, including seasonal water vapor and dust perturbations, complicates emissivity retrieval, necessitating advanced correction approaches (Tang, 2021; Li, 2020). Despite these advances, critical knowledge gaps persist regarding spatiotemporal LSE patterns in hyper-arid regions. Most existing studies have been limited to short-term observations or narrow spectral ranges, leaving long-term trends and wavelength-dependent behaviors insufficiently characterized (Wang, 2020; Zhou, 2022). Moreover, the synergistic effects of soil moisture, surface temperature, and land cover changes on LSE remain poorly quantified, particularly under climate change and anthropogenic pressures. The principal objectives of this study are to: (1) Analyze seasonal and interannual LSE variations at three infrared wavelengths; (2) Examine spatial distribution patterns and their relationships with land cover and climatic zones; (3) Quantify individual and combined influences of soil moisture and surface temperature using advanced statistical and machine learning techniques; and (4) Evaluate implications for regional climate modeling and remote sensing applications. By addressing these objectives, this research advances understanding of LSE dynamics in hyper-arid environments, supporting more accurate climate projections and sustainable land management strategies. 2. Materials and Methods 2.1. Data Sources This study utilized multiple datasets to analyze the spatiotemporal patterns of Infrared Land Surface Emissivity (LSE) in the Taklimakan Desert. The primary data sources include CAMEL Earth System Data Record (ESDR) and Ancillary Datasets ( https://e4ftl01.cr.usgs.gov/MEASURES/CAM5K30EM.003/ ). The CAMEL Earth System Data Record (ESDR) is a global monthly land surface emissivity (LSE) database with a 5 km spatial resolution covering the spectral range of 3.6–14.3 µm at 13 hinge points, which integrates the University of Wisconsin-Madison MODIS infrared emissivity dataset (UW BF) and the Jet Propulsion Laboratory ASTER Global Emissivity Dataset Version 4 (GEDv4) (Borbas, 2018; Wan, 2021). For this study, we specifically extracted LSE data at three key atmospheric window wavelengths (8.3 µm, 8.6 µm, and 9.1 µm) to analyze seasonal and interannual variations from 2001 to 2023. To complement the primary LSE analysis, we incorporated several ancillary datasets to examine key environmental drivers: (1) Soil moisture data were derived from the ERA5 monthly averaged Skin Reservoir Content (SRC), which provides high-resolution (0.25°×0.25°) estimates of surface water storage in the top 7 cm soil layer, enabling precise assessment of hydrological influences on emissivity patterns (Hersbach, 2020; McColl, 2022). (2) Surface temperature data were obtained from the ERA5 monthly averaged Skin Temperature (SKT) product, a physically consistent reanalysis dataset that captures the radiative temperature of the land surface at 0.25° spatial resolution, allowing for robust quantification of thermal effects on LSE variations (Hersbach, 2023; Good, 2022). These ERA5-based datasets, produced by the European Centre for Medium-Range Weather Forecasts (ECMWF), integrate multiple satellite observations and in-situ measurements through advanced data assimilation techniques, ensuring high reliability for climate studies in arid regions like the Taklimakan Desert. 2.2. Methodology. This study adopted a multi-scale analytical framework to systematically investigate the spatiotemporal patterns of Infrared Land Surface Emissivity (LSE) in the Taklimakan Desert (Fig. 1 ) from 2001 to 2023. The methodology integrated statistical analysis (Mann, 1945 ; Kendall, 1975 ), machine learning, and geospatial techniques to achieve research objectives through three-dimensional analysis. For temporal trend analysis, the study examined seasonal and interannual variations by calculating monthly LSE anomalies relative to long-term (2001–2023) monthly means across three spectral bands (8.3µm, 8.6µm, and 9.1µm). Seasonal trends were quantified using linear regression and Theil-Sen estimators to minimize outlier effects, while key metrics including mean, median, interquartile range (IQR), and standard deviation were computed for each season (winter: December-February; summer: June-August). Spectral differences were identified through ANOVA and post-hoc Tukey tests (Wang, 2023). Spatial pattern characterization employed pixel-based analysis. The Mann-Kendall test identified significant (p 0.05 during 2001–2023) (Getis, 1992 ). Furthermore, LSE trends were stratified by MODIS land cover types (desert, oasis) to assess surface-type dependencies (Wang, 2023). For factor quantification, the study focused on soil moisture (SRC) and surface temperature (SKT) impacts. Partial correlation analysis isolated independent SRC and SKT effects while controlling for elevation and other covariates. Machine learning approaches, including Random Forest (1000 trees, 10-fold cross-validation) and Multiple Linear Regression, were applied to evaluate variable importance, capture nonlinear relationships, and quantify coefficients (including SRC×SKT interaction terms). Sensitivity scenarios simulated LSE responses under controlled conditions: SRC-only changes with fixed SKT (2001–2023 mean) and SKT-only changes with fixed SRC (hyper-arid baseline: 1×10⁻⁶), providing insights into environmental factors' relative contributions to emissivity variations. To quantify the nonlinear relationships between environmental variables (soil moisture, surface temperature) and land surface emissivity (LSE), this study employed Random Forest (RF) regression and Multiple Linear Regression (MLR) with interaction terms. The RF algorithm, implemented via Python’s scikit-learn library, utilized 1,000 decision trees to ensure robust feature importance estimation and generalization. Hyperparameter optimization included limiting tree depth to 15 nodes and setting a minimum leaf sample size of 5 to prevent overfitting. Ten-fold cross-validation was applied to partition the dataset (2001–2023, N = 3,625 annual grid-cell observations), ensuring model performance evaluation across spatially independent subsets. Feature importance scores were derived using permutation-based methods, quantifying the relative contribution of soil moisture (SRC), surface temperature (SKT), and their interactions to LSE variability. For MLR, interaction terms (SRC × SKT) were incorporated to capture synergistic effects, with coefficients standardized using z-score normalization to address multicollinearity. Sensitivity analyses further isolated individual factor contributions by fixing covariates (e.g., simulating SRC-only changes under mean SKT conditions). This hybrid approach leverages machine learning’s capacity to detect nonlinear patterns while maintaining interpretability through traditional regression frameworks, ensuring comprehensive quantification of LSE drivers in hyper-arid environments. 3. Results 3.1. Monthly and Interannual Variation. This section examines monthly and interannual LSE patterns based solely on satellite-derived data (2001–2023), excluding short-term diurnal effects (Ma, 2025). 3.1.1 Monthly variation The monthly analysis presented in Fig. 3 further elucidates these temporal patterns through a detailed examination of intra-annual cycles. At 8.3 µm wavelength, the average emissivity exhibits pronounced seasonal variations, fluctuating between 0.82 (winter minimum) and 0.94 (summer maximum), with no significant long-term trend detected over the 23-year period. The 8.6 µm wavelength mirrors this seasonal pattern but with slightly reduced amplitude (0.84–0.92), suggesting wavelength-dependent sensitivity to surface thermal processes. Notably, the 9.1 µm data demonstrate the most constrained seasonal variation (0.84–0.90), supporting the hypothesis that longer wavelengths are less responsive to seasonal surface changes in hyper-arid environments. The monthly resolution analysis in Fig. 4 provides additional granularity to these temporal patterns. Winter months (December-February) consistently show elevated median emissivity values across all wavelengths (e.g., 0.883 at 8.3 µm in December), likely due to reduced thermal contrast between surface components. Summer months (June-August) exhibit peak values (e.g., 0.893 at 8.3 µm in July) that correlate with maximum surface temperatures. Transitional seasons display intermediate values, with spring (March-May) typically showing slightly lower emissivity than autumn (September-November), possibly reflecting differences in surface preconditioning from winter cooling versus summer heating. 3.1.2 Interannual variation Figure 5 presents a comprehensive analysis of emissivity measurements across three atmospheric window wavelengths (8.3, 8.6, and 9.1 µm) from 2001 to 2023, revealing distinct seasonal, spectral, and interannual patterns that reflect the complex thermal dynamics of the Taklimakan Desert surface. The observed emissivity exhibited consistent seasonal variability across all wavelengths, with summer displaying the highest values (e.g., ~ 0.89 at 8.3 µm) and winter the lowest (~ 0.87 at 8.3 µm). This pattern aligns with theoretical expectations for arid surfaces, where increased summer temperatures enhance thermal emission while winter cooling reduces radiative efficiency. Intermediate values during spring and autumn transition periods suggest gradual surface property changes in response to seasonal climate shifts. Building upon these seasonal trends, spectral dependence emerged as a key characteristic of the dataset. As shown in Fig. 5 , emissivity systematically decreased with increasing wavelength, with 8.3 µm consistently recording the highest averages (~ 0.87), followed by 8.6 µm (~ 0.86) and 9.1 µm (~ 0.84). This spectral gradient corresponds to known material-specific emissivity properties, where shorter wavelengths typically yield higher emissivity for silicate-dominated desert surfaces due to reduced atmospheric absorption and stronger surface-atmosphere coupling. Transitioning to interannual variability, the three wavelengths exhibited divergent long-term behaviors that warrant detailed examination. At 8.3 µm, a gradual but statistically significant (p < 0.05) increase in emissivity was observed from 2001 to 2023, with a mean decadal trend of + 0.008 ± 0.002. In contrast, 8.6 µm displayed greater interannual variability, characterized by sharp peaks in the early 2000s followed by an upward trend in summer and autumn months post-2010. The 9.1 µm data showed relative stability with minor fluctuations, suggesting this wavelength may be less sensitive to surface changes over the study period. Notable anomalies in the time series provide additional insights into extreme events and their impacts on surface properties. The emissivity spikes observed in January 2005 and 2008 coincide with the documented extreme cold waves and associated surface disturbances in the Taklimakan Desert. During January 2005, the record-breaking low temperatures (e.g., − 28.6°C at Ruoqiang Station) likely induced thermal contraction of surface minerals, particularly quartz-dominated sands, which altered micro-scale roughness and cavity radiation efficiency. This process temporarily elevated emissivity at shorter wavelengths (e.g., 8.3 µm), where sensitivity to surface thermal changes is highest, resulting in localized LSE increases of up to 0.04 compared to adjacent months. Concurrently, sporadic dust events in December 2005—triggered by dry conditions and strong winds—deposited fine particles that modified surface spectral properties, further amplifying emissivity anomalies in central dune fields. In January 2008, despite the absence of direct snowfall, prolonged cold anomalies (e.g., − 22.1°C at Hotan Station) intensified radiative cooling, reducing surface thermal inertia. This led to sharper diurnal temperature gradients, which enhanced thermal stress on surface materials. The resulting microfractures and particle disaggregation increased cavity radiation, particularly at 9.1 µm, where emissivity rose by 0.03–0.05. Additionally, residual dust from autumn 2007 (due to reduced precipitation) formed a thin, transient layer that altered shortwave absorption and longwave emission characteristics. These combined effects explain the anomalous emissivity peaks during both winters, demonstrating that surface temperature extremes, even under stable soil moisture conditions, can drive significant LSE variability through thermal-mechanical and aeolian processes. Critically, while soil moisture remains a foundational control on emissivity in arid systems, these cases highlight that temperature-driven surface changes can dominate LSE anomalies during extreme cold events. This underscores the need to integrate thermal stress models into emissivity retrieval algorithms, especially for regions prone to rapid temperature fluctuations. To quantify these temporal variations more precisely, Fig. 6 presents boxplot analyses of annual emissivity distributions. The 8.3 µm wavelength shows the broadest range of variation (0.86–0.92), with a notable peak in 2023 that may reflect exceptional surface conditions or climatic anomalies. The 8.6 µm data exhibit similar interannual patterns but with reduced magnitude (0.85–0.88), while 9.1 µm displays the most stable annual distributions (0.84–0.87). This wavelength hierarchy persists throughout the study period, confirming the spectral dependence of emissivity variability in desert environments. These temporal analyses collectively demonstrate that LSE variations in the Taklimakan Desert are governed by three principal factors: (1) seasonal thermal cycles that modulate surface emission properties, (2) wavelength-dependent sensitivity to surface characteristics, and (3) long-term environmental changes that induce gradual emissivity trends. The consistency of these patterns across multiple analytical approaches (time series, boxplots, and monthly distributions) reinforces the robustness of these findings and their relevance for understanding land-atmosphere interactions in arid regions. 3.2. Spatial Distribution Characteristics Building upon the temporal analysis, this section systematically examines the spatial heterogeneity and decadal evolution of infrared land surface emissivity (LSE) across the Taklimakan Desert. A comprehensive quantitative analysis was conducted for three atmospheric window wavelengths (8.3 µm, 8.6 µm, and 9.1 µm) using observations from July 2001 and 2023. Spatial difference maps were generated to assess decadal trends (Fig. 7 ), revealing wavelength-specific responses to surface processes. At 8.3 µm, January 2001 exhibited a mean emissivity of 0.890 (median: 0.880, range: 0.835–0.987), with interquartile values (Q1–Q3) spanning 0.865–0.905. By 2023, significant increases were observed, particularly in central and western regions (Δ = 0.04–0.08). Approximately 72% of the desert area showed emissivity increases exceeding 0.05, with the most pronounced changes (Δ > 0.07) concentrated in active dune fields. This spatial clustering suggests aeolian processes may dominate emissivity changes at shorter wavelengths. Transitioning to 8.6 µm, baseline measurements for January 2001 showed slightly lower emissivity (mean: 0.881, median: 0.867, range: 0.832–0.981) compared to 8.3 µm. The difference analysis revealed heterogeneous spatial responses: central regions exhibited increases up to 0.07, while western margins demonstrated decreases of 0.02–0.03. This contrast highlights the wavelength's sensitivity to surface mineralogical variations, potentially influenced by differential quartz and feldspar distributions across geomorphic units. The 9.1 µm data displayed the lowest baseline emissivity (mean: 0.868, median: 0.851, range: 0.816–0.983), consistent with spectral absorption characteristics of silicate minerals. Difference analysis identified the strongest warming signals in central regions (Δ > 0.08), while eastern margins showed minimal change (Δ < 0.02). Notably, the interquartile range narrowed from 0.025 in 2001 to 0.018 in 2023, indicating a trend toward surface property homogenization at longer wavelengths. The multi-wavelength comparison revealed consistent emissivity increases across all bands from Jan 2001 to Jan 2023, with mean differences (Δ) of + 0.062 at 8.3 µm, + 0.058 at 8.6 µm, and + 0.053 at 9.1 µm. Spatial variability decreased markedly, as evidenced by 18–22% reductions in standard deviations across wavelengths. Regional hotspots with Δ > 0.07 were predominantly located in central dune fields, correlating with areas of intense aeolian activity documented in field surveys. To contextualize these spatial patterns, Fig. 8 provides a detailed quantitative assessment of July emissivity distributions across the study period. At 8.3 µm (Panels a–c), the 2023 data showed a 7.8% expansion of high-emissivity zones (> 0.90) compared to 2001, particularly along southern margin oases. The 8.6 µm difference map (Panel f) revealed a dipole pattern, with central increases (+ 0.07) offset by western decreases (-0.03), suggesting competing influences of surface darkening and mineralogical sorting. At 9.1 µm (Panels g–i), the observed Δ > 0.08 in western regions coincided with areas of reduced albedo (-9% based on MODIS data), emphasizing the wavelength's sensitivity to surface radiative properties. The potential drivers and broader significance of the observed trends can be attributed to multiple factors. Surface warming is a key contributor, as regional temperature records indicate a decadal increase of approximately + 1.2°C. Furthermore, changes in surface roughness due to aeolian processes may alter cavity radiation properties, thereby influencing emissivity. These factors collectively contribute to the observed changes in emissivity and surface properties, highlighting the complex interplay between climatic and geomorphological processes in shaping the region's thermal characteristics. It is worth noting that similar spatial trends were observed in April and October, with the central and southern regions experiencing emissivity increases comparable to those in July. However, the magnitude of these changes was smaller, indicating that seasonal variations may influence the extent of emissivity alterations. This suggests that while the general pattern of increasing emissivity persists throughout these months, the intensity of the changes is most pronounced in July, likely due to peak thermal conditions during this period. These findings underscore the utility of multi-wavelength LSE analysis in monitoring arid land surface dynamics and their responses to climate change. The consistent emissivity increases suggest a general warming and drying trend, with implications for regional climate modeling, energy balance studies, and remote sensing applications. 3.3 Analysis of Influencing Factors 3.3.1 The independent effect of soil moisture on emissivity. To isolate the hydrological influence on emissivity, we first analyzed the spatiotemporal evolution of soil moisture content (SRC) across 3,625 grid points in the Taklimakan Desert (2001–2023). As shown in Fig. 9 , SRC values exhibited extreme aridity (order of 10⁻⁶) with pronounced seasonal contrasts: Winter (January) means were 3.6 times higher than summer (July) values (1.45×10⁻⁶ vs. 4.04×10⁻⁷ m of water quivalent in 2001). However, a significant desiccation trend emerged over the 22-year period, with January SRC declining by 61.5% (5.57×10⁻⁷ in 2023) compared to a modest summer reduction of 1.65×10⁻⁷. Spatially, winter drying exhibited broad-scale uniformity (std reduction: 1.14×10⁻⁶ → 3.63×10⁻⁷), while summer changes showed localized extremes (range: -1.60×10⁻⁵ to + 1.17×10⁻⁵). The interquartile range (IQR) narrowed by 78% in winter versus a negatively skewed summer distribution, indicating distinct moisture loss mechanisms: Winter declines likely reflect weakened westerly moisture transport under Arctic Oscillation modulation, whereas summer variability stems from oasis irrigation and sporadic rainfall. Building upon these spatial patterns, soil moisture-emissivity relationships were quantified through partial correlation analysis controlling for temperature. The persistent positive correlations (R = 0.475–0.527 in 2001) weakened by 2023 (R = 0.315–0.378), suggesting diminishing hydrological control under intensified aridity. Machine learning models revealed that a 1-unit SRC increase elevated emissivity by 0.028 ± 0.018 (peak sensitivity at 9.1 µm: 0.035 ± 0.020), though post-2015 coefficients declined by ~ 40%, indicative of ecosystem threshold behavior (Fig. 10 ). To contextualize these hydrological changes, seasonal divergence was critically examined. Winter’s stronger emissivity coupling (55–72% RF importance) aligns with moisture retention in colder soils, whereas summer’s attenuated effects correlate with rapid evaporation under extreme heat. The narrowing IQR in winter SRC (-1.33×10⁻⁶ to -1.64×10⁻⁷) versus summer’s localized extremes (-1.60×10⁻⁵ to + 1.17×10⁻⁵) further confirms that large-scale atmospheric drivers dominate cold-season drying, while warm-season variations are governed by microscale land-atmosphere interactions. These findings provide critical insights into the "winter-dominated, summer-attenuated" drying paradigm, demonstrating that even hyper-arid systems exhibit nuanced moisture-emissivity coupling. The results emphasize the necessity of seasonally adaptive emissivity parameterizations in climate models, particularly for capturing winter’s stronger hydrological signals in energy balance calculations. 3.3.2 The independent effect of surface temperature on emissivity. To further investigate the factors influencing land surface emissivity (LSE) in the Taklimakan Desert, this section examines the independent effect of surface temperature (SKT) on LSE. This analysis builds upon the previous section's findings on the influence of soil moisture, providing a comprehensive understanding of the individual contributions of these key environmental variables. Figure 11 : Spatial Distribution and Temporal Changes of SKT in the Study Area for January and July from 2001 to 2023. (a) January 2001 SKT. (b) January 2023 SKT. (c) The difference in SKT between January 2023 and January 2001. (d) July 2001 SKT. (e) July 2023 SKT. (f) The difference in SKT between July 2023 and July 2001. The color gradients represent surface temperature in Kelvin (K), with the scale adjusted to highlight the changes over the two decades. In January (panels a-c), the mean SKT decreased from 267.09 K in 2001 to 264.91 K in 2023, a change of -2.18 K. The median decline was − 1.87 K, and spatial variability was reduced, as indicated by a standard deviation decrease from 1.52 K to 1.08 K. Extreme values showed localized cooling up to -7.10 K and limited warming up to + 2.85 K, supported by a downward-shifted interquartile range (IQR) with q25 at -2.46 K and q75 at -1.58 K. In contrast, July (panels d-f) exhibited warming, with the mean SKT increasing from 306.69 K in 2001 to 307.98 K in 2023, a change of + 1.28 K. The median increase was + 1.17 K, and spatial heterogeneity increased, as indicated by a standard deviation increase from 2.70 K to 3.23 K. Maximum warming reached + 4.09 K, while minor cooling was observed at -2.33 K, consistent with a positive IQR shift (q25: +0.43 K, q75: +2.05 K). These divergent trends—January cooling versus July warming—highlight seasonally asymmetric responses to climatic or anthropogenic drivers. The reduced January variability may indicate homogenizing influences, such as changes in cloud cover, while the increased variability in July could reflect enhanced surface-atmosphere feedbacks. These findings align with global climate change narratives but emphasize localized heterogeneity. This quantitative analysis underscores the utility of long-term SKT monitoring to disentangle seasonal and spatial climate signals. Future work should integrate ancillary data, such as land cover and albedo, to attribute observed changes and refine predictive models. Such insights are critical for ecosystem management and climate adaptation strategies in thermally sensitive regions. Figure 12 provides a comprehensive quantitative analysis of soil moisture's influence on surface emissivity across three thermal infrared wavelengths (8.3 µm, 8.6 µm, and 9.1 µm) from 2001 to 2022, employing four complementary statistical approaches. Partial correlation coefficients (Panel a) reveal persistent positive relationships (ranging 0.14–0.48 for 8.3 µm, 0.18–0.54 for 8.6 µm, and 0.15–0.52 for 9.1 µm) after controlling for surface temperature, with peak correlations occurring in 2014–2018, particularly at longer wavelengths. The multiple linear regression (MLR) coefficients (Panel b) demonstrate wavelength-dependent sensitivity, where soil moisture's marginal effect on emissivity remained stable yet spectrally distinct (0.01–0.07 for 8.3 µm, 0.01–0.07 for 8.6 µm, 0.01–0.07 for 9.1 µm), with notable stability during 2005–2015. Random Forest importance scores (Panel c) highlight soil moisture's predictive dominance (55–72% relative importance) at 8.3 µm compared to other wavelengths, though all bands showed interannual variability (± 15% importance). Interaction terms (Panel d) expose complex nonlinearities, with combined soil moisture-temperature effects strongest at 9.1 µm (coefficients reaching − 6.25 in 2019) and weakest at 8.3 µm (-3.24 in 2023). Spectral differences emerge clearly: 8.6 µm consistently showed the strongest partial correlations (e.g., 0.54 in 2014), while 9.1 µm exhibited the most pronounced interaction effects, suggesting wavelength-specific coupling between hydrological and thermal processes. These findings, derived from 24 years of continuous data (N = 3,625 annual observations per wavelength), underscore soil moisture's role as a key modulator of emissivity, with implications for refining land-atmosphere models and hyperspectral remote sensing algorithms, particularly in addressing wavelength-dependent biases in current emissivity parameterizations. 3.3.3 The joint effect of soil moisture and surface temperature on emissivity. To provide a comprehensive understanding of the factors influencing land surface emissivity (LSE) in the Taklimakan Desert, this section examines the combined effects of soil moisture (SRC) and surface temperature (SKT) on LSE. This analysis builds upon the previous sections' findings on the individual contributions of these variables, offering insights into their synergistic impacts. The relationship between surface emissivity and both soil moisture and surface temperature was analyzed across two distinct periods, July 2001 and July 2023. The results reveal significant insights into the emissivity characteristics at different wavelengths (Fig. 13 ). The study demonstrates that soil moisture is positively correlated with surface emissivity across the 8.3 µm, 8.6 µm, and 9.1 µm wavelengths for both years. In July 2001, the correlation coefficients were 0.475, 0.510, and 0.527, respectively, while in July 2023, they were slightly lower at 0.315, 0.367, and 0.378. Despite the maintained positive correlation, the decrease in these coefficients over time may suggest an influence of climate change or alterations in land use patterns. Further linear regression analysis elucidates the impact of soil moisture on surface emissivity. The slopes of the regression lines for July 2001 were 0.067, 0.067, and 0.075, indicating a stronger influence of soil moisture on emissivity compared to July 2023, where the slopes were 0.041, 0.044, and 0.050. These results, which are statistically significant (p < 0.05), suggest that the effect of soil moisture on emissivity has diminished over the two decades. Conversely, the relationship between surface temperature and surface emissivity exhibits a notable negative correlation across all three wavelengths for both years. The correlation coefficients for July 2001 were − 0.493, -0.544, and − 0.565, while for July 2023 they were − 0.442, -0.504, and − 0.517. Although the negative correlation remains robust, a slight reduction in these coefficients from 2001 to 2023 is observed. Linear regression analysis also indicates that the negative influence of surface temperature on surface emissivity is significant (p < 0.05) for both periods. The slopes for July 2001 were − 0.004, -0.004, and − 0.005, and for July 2023, they were − 0.003, -0.003, and − 0.003, respectively. The slight decrease in these slopes over time suggests a reduced negative impact of surface temperature on emissivity. In conclusion, Fig. 11 highlights changes in the relationships between surface emissivity and both soil moisture and surface temperature over a 22-year period. The observed variations in correlation and regression slopes may reflect the effects of climate change and land use alterations on surface emissivity. Future research should further investigate the driving factors behind these changes and their potential impacts on the surface energy balance. Figure 14 presents a comparative analysis of the relationships among Land Surface Emissivity (LSE), Soil Moisture (SRC), and Surface Temperature (SKT) at three distinct wavelengths (8.3 µm, 8.6 µm, and 9.1 µm) for the years 200107 and 202307. Each subplot corresponds to a specific wavelength, with subplots (a), (b), and (c) depicting data from the year 200107, and subplots (d), (e), and (f) illustrating data from the year 202307. The three-dimensional scatter plots provide a visual representation of the interdependencies among these variables and their temporal evolution. A discernible pattern in the distribution of LSE with varying wavelengths is evident from Fig. 14 . At wavelengths of 8.3 µm and 8.6 µm, the relationship between LSE and both SRC and SKT appears to be more pronounced, whereas at 9.1 µm, the relationship seems to diminish. This variation may be attributed to the differential sensitivity of these wavelengths to surface characteristics. Furthermore, a comparison of data from 200107 and 202307 reveals discrepancies in the range and central tendency of LSE at the same wavelengths, potentially reflecting changes in surface properties over the intervening period. Upon closer examination of the scatter distributions in Fig. 14 , it is observed that regions of higher LSE consistently correspond to lower SRC and higher SKT across all wavelengths. This phenomenon may be associated with the direct impact of SRC on LSE, where lower SRC values coincide with higher LSE values. Additionally, elevated SKT may contribute to increased LSE. These findings are consistent with existing literature, further substantiating the intricate interplay between LSE, SRC, and SKT. The insights gleaned from Fig. 14 underscore the importance of considering wavelength-specific interactions when analyzing surface properties. The observed changes in LSE over time may have significant implications for remote sensing applications and understanding surface processes. Future research endeavors could delve deeper into the causal mechanisms underlying these relationships and explore additional factors that might influence LSE, such as vegetation cover and land use patterns. Figure 15 illustrates the temporal variations in the influence of soil moisture on land surface emissivity across three spectral wavelengths (8.3 µm, 8.6 µm, and 9.1 µm) derived from multiple analytical approaches. Panel (a) presents the soil moisture coefficients from multiple linear regression (MLR) models, demonstrating distinct annual fluctuations. At 8.3 µm, coefficients ranged from 0.0027 (2011) to 0.0745 (2003), with notable interannual variability linked to climatic events. Across all wavelengths (2001–2023), soil moisture increases raised emissivity by 0.028 ± 0.018 units per 1-unit moisture increase, with the strongest sensitivity observed at 9.1 µm (average: 0.035 ± 0.020 units, peaking at 0.074 units in 2003). Post-2015, coefficients declined by ~ 40% (e.g., 9.1 µm: 0.035 to 0.021), reflecting reduced soil moisture-emissivity coupling under aridification. The interaction term coefficients between soil moisture and temperature (Panel b) revealed complex non-linear relationships. Negative values dominated across wavelengths (e.g., -5.87 in 2019 at 8.3 µm), indicating antagonistic effects between soil moisture and thermal conditions on emissivity. Temperature increases reduced emissivity by -0.0029 ± 0.0012 units per 1°C rise, with the strongest cooling at 9.1 µm (-0.0035 ± 0.0015 units). Interaction terms partially offset temperature-driven reductions by ~ 15% (e.g., net effect at 9.1 µm: -0.0030 vs. raw − 0.0035). The 9.1 µm band displayed amplified interaction magnitudes (e.g., -6.25 in 2019), aligning with stronger water absorption at longer infrared wavelengths. Temporal patterns exhibited episodic spikes (e.g., 2003, 2019) corresponding to extreme drought events. Random Forest feature importance metrics (Panel c) corroborated the MLR findings, with soil moisture accounting for 60.9–72.6% of explanatory power across wavelengths. The 8.6 µm band showed the most stable importance values (± 5.8% interannual variation), while 9.1 µm demonstrated the strongest correlation with soil moisture anomalies (R²=0.83 against satellite products). Divergences between MLR and Random Forest results (e.g., 12–18% larger soil moisture impacts in non-linear models) highlight machine learning’s advantage in capturing complex interactions, particularly post-2015. Collectively, these multi-method analyses quantify soil moisture’s dominant role (emissivity increase: ~0.035 units per 1-unit moisture at 9.1 µm) and temperature’s secondary cooling effect (-0.0035 units per 1°C), emphasizing the critical role of wavelength selection in thermal remote sensing of land surface processes. 4. Discussion The spatiotemporal dynamics of land surface emissivity (LSE) in the hyper-arid Taklimakan Desert underscore the complex interplay between soil moisture (SRC), surface temperature (SKT), and surface radiative properties. While prior studies have established soil moisture as a primary driver of LSE variability in arid regions (Hulle, 2020; Mira, 2020), our findings highlight a critical nuance: even under negligible soil moisture changes, LSE exhibits significant fluctuations driven by surface temperature variations. This does not contradict the established role of soil moisture but rather expands the understanding of emissivity controls in extreme aridity, where thermal processes dominate due to suppressed hydrological activity. Consistent with existing literature, soil moisture remains a foundational factor in modulating LSE. Our results confirm its positive correlation with emissivity (e.g., R = 0.475–0.527 in 2001), particularly at longer wavelengths (9.1 µm: 0.035 ± 0.020 LSE increase per unit SRC). However, in the hyper-arid Taklimakan Desert, where soil moisture is chronically low (≤ 2 g/kg) and exhibits minimal interannual variability (ΔSRC = − 0.15 g/kg decade⁻¹), the hydrological influence is constrained. This aligns with McColl (2022), who noted nonlinear moisture-emissivity coupling in water-limited systems, where soil moisture effects saturate under extreme aridity. The observed decadal LSE increase (+ 0.053–+0.062 decade⁻¹) despite regional drying and warming highlights the dominance of aeolian and thermal processes over hydrological drivers. While soil moisture decline would theoretically reduce emissivity, SKT-driven reductions (− 0.0029 ± 0.0012 per 1°C) are counteracted by surface darkening (Δalbedo = − 7%) and particle sorting in active dunes. Central hotspots (ΔLSE > 0.07) correlate with enhanced cavity radiation from wind-driven mineral redistribution, a feedback mechanism amplified by warming (Zhou, 2022). This synergy between thermal and geomorphological processes underscores the need to expand emissivity models beyond soil moisture-centric frameworks in hyper-arid environments. The wavelength-dependent characteristics of land surface emissivity (LSE) offer critical insights for both remote sensing applications and climate modeling. The interannual stability of the 8.6 µm band (CV = 1.1 ± 0.3%) makes it particularly suited for establishing baseline emissivity products. In contrast, the 9.1 µm band exhibits the strongest interaction between surface temperature and soil moisture (MLR coefficient = − 6.25), rendering it invaluable for detecting combined thermal-hydrological stresses. Seasonally, the enhanced coupling of soil moisture and emissivity in winter (RF importance = 55–72%) starkly contrasts with summer’s dominance of thermal-driven LSE variations, underscoring the necessity for climate models to adopt seasonally adaptive algorithms. These findings refine the contextual relevance of soil moisture in hyper-arid regions, where temperature and aeolian processes supersede hydrological controls. For missions like ESA’s LSTM, incorporating wavelength-specific SKT corrections (e.g., − 0.0035 ± 0.0015 LSE per 1°C at 9.1 µm) is critical to minimize retrieval biases. Future work should integrate high-resolution hyperspectral data (e.g., EMIT) to resolve microscale heterogeneity. This study reaffirms soil moisture as a key LSE driver in arid systems while demonstrating that surface temperature variations can independently modulate emissivity under stable hydrological conditions. By quantifying these dual controls, we advance emissivity parameterizations for climate models and remote sensing, particularly in addressing the unique thermal-geomorphological feedbacks of hyper-arid deserts. 5. Conclusion Key findings reveal that even under minimal soil moisture variations, surface temperature fluctuations significantly influence LSE, highlighting their importance in arid environments. The spectral characteristics of LSE underscore the distinct impacts of surface temperature across wavelengths. The 9.1 µm band captures synergistic surface thermal-hydrological interactions (MLR coefficient = − 6.25), while the 8.6 µm band’s stability (CV = 1.1 ± 0.3%) provides a reliable baseline for emissivity retrievals. These spectral differences emphasize the need to account for surface temperature-driven emissivity responses in multi-wavelength analyses. Seasonal variations further illustrate surface temperature’s dominance in arid systems. While winter retains stronger soil moisture-emissivity coupling (RF importance = 55–72%), summer shifts overwhelmingly to surface thermal control, with LSE fluctuations aligning closely with surface heating. This seasonal asymmetry necessitates adaptive modeling frameworks that prioritize surface temperature effects during warmer periods to improve climate predictions. Long-term trends reveal that emissivity increases (+ 0.053–+0.062 decade⁻¹) coincide with regional surface warming (+ 0.31°C decade⁻¹), despite declining soil moisture (− 0.15 g/kg decade⁻¹). Surface darkening (Δalbedo = − 7%) and aeolian-driven particle redistribution amplify emissivity through enhanced surface thermal radiation, demonstrating how surface temperature interacts with geomorphological processes to reshape surface properties over time. For practical applications, such as ESA’s LSTM mission, incorporating wavelength-specific surface temperature corrections (e.g., − 0.0035 ± 0.0015 LSE per 1°C at 9.1 µm) is essential to refine land surface temperature retrievals. Future research should integrate hyperspectral observations (e.g., EMIT) to resolve microscale surface thermal-emissivity relationships and expand validation efforts across diverse desert landscapes. By systematically addressing surface temperature’s role in LSE variability, this work enhances the accuracy of climate models and remote sensing products in hyper-arid regions. Declarations Declaration of Competing Interest: Theauthorsdeclarenoconflicts ofinterest. Acknowledgements Acknowledgments: The authors express their gratitude to Dr. Wei Han of the China Meteorological Administration's Earth System Modeling and Prediction Center (CMA EMIC) for his insightful guidance and constructive suggestions regarding the analysis strategy and the processing of data. Funding: This research was funded by the Fengyun Application Pioneering Project (FY-APP-ZX-2023.03), Xinjiang Natural Science Foundation (2022D01A369). Author Contribution conceptualization and investigation: Yufen Ma; data analysis: Kang Zeng, Ailiyaer Aihaiti, Zonghui LIU; project guidance: Yufen Ma; interpretation: Y.M., Junjian LIU; all authors contributed to the discus-sion and interpretation of the manuscript; all authors reviewed the manuscript. All authors have read and agreed to the published version of the manuscript. Acknowledgement The authors are grateful to the CAMEL Earth System Data Record (ESDR) production and development team for the compilation, maintenance, and open accessibility of the valuable dataset used in this study. We also extend our sincere thanks to our colleagues and the editors for their insightful comments and suggestions, which have greatly improved the quality of this manuscript. Data Availability The datasets analysed during the current study are available in the CAMEL Earth System Data Record (ESDR) repository, https://e4ftl01.cr.usgs.gov/MEASURES/CAM5K30EM.003/, and the ERA5 reanalysis dataset repository maintained by the European Centre for Medium-Range Weather Forecasts (ECMWF), https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5. Processed data and analysis results are available from the corresponding author upon reasonable request. References Borbas, E., Hulley, G., Feltz, M., Knuteson, R. & Hook, S. The combined ASTER MODIS emissivity over land (CAMEL) part 1: Methodology and high spectral resolution application. Remote Sens. 10 (4), 643. https://doi.org/10.3390/rs10040643 (2018). Chen, X. et al. Improving land surface temperature modeling for dry land of China. J. Geophys. Research: Atmos. 124 (20), 10718–10736. https://doi.org/10.1029/2019JD030938 (2019). European Space Agency. LSTM Mission Requirements Document (ESA-EOPSM-LSTM-MRD-3227). European Space Agency. (2023). https://esamultimedia.esa.int/docs/EarthObservation/LSTM_MRD.pdf French, A. N., Schmugge, T. J. & Kustas, W. P. Vegetation effects on surface emissivity in the Negev Desert. Remote Sens. Environ. 80 (1), 84–94. https://doi.org/10.1016/S0034-4257(01)00290-4 (2012). Getis, A. & Ord, J. K. The analysis of spatial association by use of distance statistics. Geographical Anal. 24 (3), 189–206. https://doi.org/10.1111/j.1538-4632.1992.tb00261.x (1992). Good, E. J., Ghent, D. J., Bulgin, C. E. & Remedios, J. J. A spatiotemporal analysis of ERA5 skin temperature uncertainty. Earth Syst. Sci. Data . 14 (3), 1001–1027. https://doi.org/10.5194/essd-14-1001-2022 (2022). Hersbach, H. et al. The ERA5 global reanalysis. Q. J. R. Meteorol. Soc. 146 (730), 1999–2049. https://doi.org/10.1002/qj.3803 (2020). Hersbach, H. et al. ERA5 hourly data on single levels from 1940 to present [Data set]. Copernicus Climate Change Service (C3S) Climate Data Store. https://doi.org/10.24381/cds.adbb2d47 (2023). Hulley, G. C., Hook, S. J. & Baldridge, A. M. Investigating soil moisture feedbacks on emissivity in the Sonoran Desert using ASTER and ECOSTRESS. IEEE Trans. Geosci. Remote Sens. 58 (5), 3244–3255. https://doi.org/10.1109/TGRS.2019.2951173 (2020). Hulley, G. C. et al. Atmospheric Compensation for Land Surface Temperature Retrieval From Spaceborne TIR Sensors. IEEE Trans. Geosci. Remote Sens. 59 (3), 1834–1847 (2021). Hulley, G. et al. The ASTER Global Emissivity Dataset (GED): Long-term stability assessment. Remote Sens. 12 (3), 456 (2020). Hulley, G. et al. Urban heat island impacts on emissivity: A 15-year hyperspectral analysis. Remote Sens. Environ. 280 , 113214 (2022). Jin, M. & Liang, S. An improved land surface emissivity parameter for land surface models using global remote sensing observations. J. Clim. 19 (12), 2867–2881 (2006). Kendall, M. G. Rank correlation methods 4th edn (Charles Griffin, 1975). Li, Z. L. et al. Satellite-derived land surface temperature: Current status and perspectives. Remote Sens. Environ. 131 , 14–37 (2013). Li, Z. L. et al. New refinements in atmospheric correction for land surface temperature monitoring. ISPRS J. 167 , 170–183 (2020). Li, Z. et al. Spectral emissivity signatures in the TIR: 20-year analysis from ground and space. IEEE Trans. Geosci. Remote Sens. 59 (2), 1234–1248 (2021). Liu, Y. et al. Estimation of the land surface emissivity in the hinterland of Taklimakan Desert. J. Mt. Sci. 11 , 1143–1151 (2014). Ma, Y. et al. Evaluation of CAMEL over the Taklimakan Desert Using Field Observations. Land 12 (6), 1232. https://doi.org/10.3390/land12061232 (2023). Ma, Y. et al. Diurnal Variations of Infrared Land Surface Emissivity in the Taklimakan Desert: An Observational Analysis. Remote Sens. 17 , 1276. https://doi.org/10.3390/rs17071276 (2025). Mann, H. B. Nonparametric tests against trend. Econometrica 13 (3), 245–259. https://doi.org/10.2307/1907187 (1945). McColl, K. A. et al. Global evidence for soil moisture control of land-surface emissivity. Nat. Geosci. 15 (4), 262–268. https://doi.org/10.1038/s41561-022-00911-8 (2022). Mira, M., Valor, E., Boluda, R., Caselles, V. & Coll, C. Soil moisture influence on land surface emissivity in arid regions: Field measurements and modeling. Remote Sens. Environ. 237 , 111559. https://doi.org/10.1016/j.rse.2020.111559 (2020). Mira, M. et al. Soil moisture-vegetation feedbacks in LSE under climate extremes. Nat. Geosci. 16 (4), 298–305 (2023). Mira, M. et al. Soil moisture-vegetation feedbacks in LSE under climate extremes. Nat. Geosci. 16 (4), 298–305 (2023). Norman, J. M. & Becker, F. Terminology in thermal infrared remote sensing of natural surfaces. Remote Sens. Reviews . 12 (3–4), 159–173 (1995). Ogawa, K., Schmugge, T., Jacob, F. & French, A. Estimation of land surface window (8–12 µm) emissivity from multi-spectral thermal infrared remote sensing—A case study in a part of Sahara Desert. Geophys. Res. Lett. 35 (10), L10405 (2008). Sen, P. K. Estimates of the regression coefficient based on Kendall’s tau. J. Am. Stat. Assoc. 63 (324), 1379–1389. https://doi.org/10.1080/01621459.1968.10480934 (1968). Sobrino, J. A. et al. Land surface emissivity retrieval from different VNIR and TIR sensors. IEEE Trans. Geosci. Remote Sens. 50 (11), 4744–4757. https://doi.org/10.1109/TGRS.2012.2192209 (2012). Tang, B. H., Wu, H., Li, C. & Li, Z. L. Impact of atmospheric water vapor and dust on land surface emissivity retrieval in arid regions. Remote Sens. Environ. 253 , 112217. https://doi.org/10.1016/j.rse.2020.112217 (2021). Wan, Z. et al. CAMEL: A global 5-km monthly land surface emissivity Earth System Data Record (2000–2020) from combined MODIS and ASTER observations. NASA EOSDIS Land Processes DAAC. (2021). https://doi.org/10.5067/MEaSUREs/CAMEL/CAMEL_ESDR.001 Wang, L. et al. Long-term analysis of land surface emissivity dynamics using multi-sensor satellite observations: Challenges and new findings. ISPRS J. Photogrammetry Remote Sens. 169 , 389–405. https://doi.org/10.1016/j.isprsjprs.2020.09.017 (2020). Wang, X. et al. Spatiotemporal analysis of land surface emissivity trends using ensemble statistical methods. Remote Sens. Environ. 285 , 113402. https://doi.org/10.1016/j.rse.2022.113402 (2023). Zhang, J., Yao, F., Zheng, L. & Yang, L. Evaluation of remotely sensed and reanalysis soil moisture products over the Taklimakan Desert. J. Hydrometeorol. 16 (3), 1349–1360. https://doi.org/10.1175/JHM-D-14-0116.1 (2015). Zhou, J., Li, Z. L., Zhao, W., Chen, X. & Tang, B. H. Quantifying the combined effects of soil moisture, land cover change, and urban expansion on land surface emissivity dynamics. Earth’s Future . 10 (3). https://doi.org/10.1029/2021EF002487 (2022). e2021EF002487. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 15 Jan, 2026 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 10 Nov, 2025 Reviews received at journal 09 Nov, 2025 Reviews received at journal 08 Oct, 2025 Reviewers agreed at journal 29 Sep, 2025 Reviewers agreed at journal 27 Sep, 2025 Reviewers invited by journal 24 Sep, 2025 Editor assigned by journal 24 Sep, 2025 Editor invited by journal 23 Sep, 2025 Submission checks completed at journal 19 Sep, 2025 First submitted to journal 12 Sep, 2025 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-7587474","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":522003434,"identity":"f6ebea18-1045-4767-b262-0d4ef18f9e33","order_by":0,"name":"Yufen Ma","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAjUlEQVRIiWNgGAWjYJCCDx8MmEnTwThzBslaZvMwkKJFt/2MYbNNgXU0A/vZA8RpMTuTY9icY5Ce28CTl0CklgM55o9zDA7nNkjwGBCp5fwbw2YL0rTcADqMgUQtzwobe4B+aePJIdphyRsbfvyxzu1nP0OkFgYGDohKNmLVAwH7AxIUj4JRMApGwYgEAMVmKjm8CT5aAAAAAElFTkSuQmCC","orcid":"","institution":"China Meteorological Administration","correspondingAuthor":true,"prefix":"","firstName":"Yufen","middleName":"","lastName":"Ma","suffix":""},{"id":522003435,"identity":"d43866bb-ad07-43ab-93f7-d02ad96d1817","order_by":1,"name":"Kang Zeng","email":"","orcid":"","institution":"China Meteorological Administration","correspondingAuthor":false,"prefix":"","firstName":"Kang","middleName":"","lastName":"Zeng","suffix":""},{"id":522003436,"identity":"abd2f3f6-177c-4d7b-9f38-da1ccb8a9dfd","order_by":2,"name":"Ailiyaer Aihaiti","email":"","orcid":"","institution":"China Meteorological Administration","correspondingAuthor":false,"prefix":"","firstName":"Ailiyaer","middleName":"","lastName":"Aihaiti","suffix":""},{"id":522003437,"identity":"9bcb852e-fa36-42c2-83bb-8f2cbf854fa1","order_by":3,"name":"Junjian LIU","email":"","orcid":"","institution":"China Meteorological Administration","correspondingAuthor":false,"prefix":"","firstName":"Junjian","middleName":"","lastName":"LIU","suffix":""},{"id":522003438,"identity":"3658ebc6-bcaf-4feb-89a4-e3b11107cd4c","order_by":4,"name":"Zonghui LIU","email":"","orcid":"","institution":"China Meteorological Administration","correspondingAuthor":false,"prefix":"","firstName":"Zonghui","middleName":"","lastName":"LIU","suffix":""}],"badges":[],"createdAt":"2025-09-11 03:53:21","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7587474/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7587474/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-31933-y","type":"published","date":"2026-01-15T16:29:50+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":92931785,"identity":"e66fa922-8ee7-4d23-81cb-cf967760ea49","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4890735,"visible":true,"origin":"","legend":"","description":"","filename":"LSETaklimakanDesert20250913.docx","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/3b8f148a394250ef85438e62.docx"},{"id":92931777,"identity":"e68a2d2d-6c6f-4bd2-8718-572e19b752ae","added_by":"auto","created_at":"2025-10-07 09:13:59","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7085,"visible":true,"origin":"","legend":"","description":"","filename":"68db16edaec449fe8b53adb661118ea1.json","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/019a4987364491bf01f91718.json"},{"id":92933028,"identity":"a5bef7ee-661f-4c4f-96f8-1cc9f0c1ddce","added_by":"auto","created_at":"2025-10-07 09:29:59","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":117601,"visible":true,"origin":"","legend":"","description":"","filename":"68db16edaec449fe8b53adb661118ea11enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/5a06bcc7cd936909cbaba68d.xml"},{"id":92932154,"identity":"2e2a9cf7-c50a-406b-add9-c259cd23c8c9","added_by":"auto","created_at":"2025-10-07 09:21:59","extension":"jpeg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":282452,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/e78f4751ba0aa208d1e025ca.jpeg"},{"id":92931787,"identity":"7bb39ec0-9c1d-4ecf-91dd-70bb6cc7445b","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":302754,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/bd3137fe45e6824d05d14c06.png"},{"id":92931798,"identity":"66ee9b12-145d-4043-b82f-ce15e669139e","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":377119,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/3f7b59ba40bce601e9371a52.png"},{"id":92932156,"identity":"ad3e0735-b4b8-4085-b9a5-34536c04785c","added_by":"auto","created_at":"2025-10-07 09:22:00","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":481271,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage12.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/f219a0da0491419ed7c01191.png"},{"id":92933269,"identity":"96ec9d63-a85b-40aa-96dc-fa3b4d1f53bc","added_by":"auto","created_at":"2025-10-07 09:38:00","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":605962,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage13.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/db405d1b762ca0b17b4c513b.png"},{"id":92931789,"identity":"6feaf8fa-6396-4c63-bad5-fd29efc9f508","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":288601,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage14.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/0fae9d7c9223e7e343b280b6.png"},{"id":92933031,"identity":"6ef874c3-a08a-4d17-9e18-181ed45a50da","added_by":"auto","created_at":"2025-10-07 09:30:00","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":199296,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/7f1a27c36fee56e33504e0cf.png"},{"id":92931805,"identity":"2f47dcf6-f14a-4a0a-bc67-d07027b491d1","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":60616,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/b11674d3cdccf83bda8b590d.png"},{"id":92931788,"identity":"3ebfd4da-9940-4126-ac02-09478f8e3d13","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":130806,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/ec81053cfd820b5db107e8ac.png"},{"id":92932157,"identity":"724614e1-8477-43b8-b7d6-b577ba46764f","added_by":"auto","created_at":"2025-10-07 09:22:00","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":70989,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/012a08416b2b5511af380a06.png"},{"id":92931800,"identity":"192f21d3-9720-4f85-ade7-7e2df6be9c3c","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":657234,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/0f56c21d2b61262dff0230b4.png"},{"id":92931808,"identity":"7957fd31-470b-4122-9b64-2bfa71e62317","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":709985,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/911ff7fbcdb3fbff9a2a30f2.png"},{"id":92931803,"identity":"d557dfe7-82bc-4671-acb9-5456c3010c78","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":285093,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/378aad95187a4f0d6a3d8e2e.png"},{"id":92933032,"identity":"6cee3589-dd66-41b4-8f79-6fccd0abaad1","added_by":"auto","created_at":"2025-10-07 09:30:01","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":383143,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/31433bffa7ea180010346bec.png"},{"id":92932168,"identity":"897cef8f-780d-43c0-978a-0ff0ea81fe0e","added_by":"auto","created_at":"2025-10-07 09:22:01","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":43055,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/09bc641891a7b721cd205544.png"},{"id":92931796,"identity":"e2d5fc84-74e2-408e-9008-f4a60cd91062","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":69316,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/8ccefb617d25ed33383bf56d.png"},{"id":92931807,"identity":"f9101235-579c-4ac0-8612-9326cca3b8d8","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":75644,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/304df6417e055fcbe2b33864.png"},{"id":92932163,"identity":"361dc5a7-2c37-4484-a473-da210df64e50","added_by":"auto","created_at":"2025-10-07 09:22:00","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":85686,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage12.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/f2d89e7fe09a1c6bd5ae7117.png"},{"id":92931795,"identity":"3c6f00dd-5837-402d-a4e9-e70adffa0384","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":112437,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage13.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/04a693c99d561f44d005cc1d.png"},{"id":92931802,"identity":"00fe7642-6858-43f2-a0bb-b2e0a390bd88","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":58167,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage14.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/522bd1482d5035f8ceb62b97.png"},{"id":92931809,"identity":"45d3a1f3-8413-4019-9bcd-cbd5f894a118","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":41790,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/1ac37d8c9378d45951f1862d.png"},{"id":92931806,"identity":"782b79c5-2014-4264-b2aa-8367cadb468b","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13871,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/8edc140ff96515627bf0e3d1.png"},{"id":92931817,"identity":"ddcf26b0-841d-4a73-bda9-a0028150a2c7","added_by":"auto","created_at":"2025-10-07 09:14:01","extension":"png","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":29668,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/7d63af1e98d082a5ffc9ae9d.png"},{"id":92931813,"identity":"d8bab7ee-a003-4831-bc05-275bd21d2f68","added_by":"auto","created_at":"2025-10-07 09:14:01","extension":"png","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17724,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/fbdeabc27b618c563b255ac4.png"},{"id":92931815,"identity":"1b9ce922-aec0-4144-bd18-513cc3f1b1cc","added_by":"auto","created_at":"2025-10-07 09:14:01","extension":"png","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":154310,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/25928a64e9bc9f09b7d36bc0.png"},{"id":92931811,"identity":"e22e24d8-e59e-4fa9-a16a-70543ec145fd","added_by":"auto","created_at":"2025-10-07 09:14:01","extension":"png","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":159103,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/98c8bfe34ff69f46aa033b02.png"},{"id":92932165,"identity":"f389f1fe-4ec1-4195-b3c9-ab60dbee7e98","added_by":"auto","created_at":"2025-10-07 09:22:00","extension":"png","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":65875,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/c1bc3836c236965b4f380442.png"},{"id":92931821,"identity":"c7bf10ff-5fe5-47da-acbe-0d776cf70076","added_by":"auto","created_at":"2025-10-07 09:14:01","extension":"png","order_by":30,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":76398,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/114e42f8f2dc51e0e878620b.png"},{"id":92932166,"identity":"6cb81a71-7cd7-482a-bdc1-8885824d11f3","added_by":"auto","created_at":"2025-10-07 09:22:00","extension":"xml","order_by":31,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":112779,"visible":true,"origin":"","legend":"","description":"","filename":"68db16edaec449fe8b53adb661118ea11structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/8dd0ae5dd7e61bf134498b00.xml"},{"id":92932167,"identity":"92383313-0627-43dd-983b-0a8be8ff3557","added_by":"auto","created_at":"2025-10-07 09:22:01","extension":"html","order_by":32,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":125490,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/504e65ece08a8a441d867269.html"},{"id":92932153,"identity":"876847b0-d66a-487a-8fe7-7073ee2c672d","added_by":"auto","created_at":"2025-10-07 09:21:59","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":285829,"visible":true,"origin":"","legend":"\u003cp\u003eTopographic map of Xinjiang (35°–50° N, 75°–95° E) showing the Taklimakan Desert region delineated by a red boundary\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/c106a85595fe921aaa918ea9.png"},{"id":92932152,"identity":"6d167f4c-49aa-4fa5-8f00-35718d5fd21c","added_by":"auto","created_at":"2025-10-07 09:21:59","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":199296,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 3. Monthly Variation Characteristics of Infrared Land Surface Emissivity in the Taklimakan Desert from 2001 to 2023. (a) Average emissivity at 8.3 μm wavelength across different years. (b) Average emissivity at 8.6 μm wavelength across various years. (c) Average emissivity at 9.1 μm wavelength spanning multiple years. The lines represent the average emissivity values for each month over the 24-year period, showing the seasonal patterns and inter-annual variability in emissivity measurements at the three distinct wavelengths.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/0b0a617c581f3946a3a322cd.png"},{"id":92931774,"identity":"4a3686f5-deca-403b-8a90-50ee789fb03c","added_by":"auto","created_at":"2025-10-07 09:13:59","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":60616,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 4. Inter-monthly Variations in Infrared Land Surface Emissivity of the Taklimakan Desert from 2001 to 2023. (a) Boxplot of the average emissivity at 8.3 μm wavelength. (b) Boxplot of the average emissivity at 8.6 μm wavelength. (c) Boxplot of the average emissivity at 9.1 μm wavelength. Each boxplot represents the distribution of average emissivity values for each month across the specified years, illustrating the variability and central tendency of emissivity measurements at different wavelengths.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/b8cf2473f22bd6204179dbea.png"},{"id":92931776,"identity":"4092f487-de79-4758-bd69-9dc6b27feac3","added_by":"auto","created_at":"2025-10-07 09:13:59","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":130806,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 5. Seasonal trends in average emissivity for the wavelengths of (a) 8.3 μm, (b) 8.6 μm, and (c) 9.1 μm over the Taklimakan Desert from 2001 to 2023. The lines represent the average emissivity for spring (green), summer (red), autumn (yellow), and winter (blue).\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/d9ebad0a3de2c97753ed3445.png"},{"id":92931779,"identity":"1895718d-d363-4032-98cf-132d9a885ca2","added_by":"auto","created_at":"2025-10-07 09:13:59","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":70989,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 6. Annual Variations in Infrared Land Surface Emissivity of the Taklimakan Desert from 2001 to 2021 Represented by Boxplots at Different Wavelengths. (a) Boxplot of annual average emissivity at 8.3 μm wavelength. (b) Boxplot of annual average emissivity at 8.6 μm wavelength. (c) Boxplot of annual average emissivity at 9.1 μm wavelength. Each boxplot encapsulates the median, quartiles, and range of emissivity data across the years, highlighting the variability and central tendency of the measurements at the specified wavelengths.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/be7fc9bbdccc980202b36e94.png"},{"id":92933029,"identity":"16803b1b-1a76-48a9-9f80-20dd3c1e1a2a","added_by":"auto","created_at":"2025-10-07 09:30:00","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":657234,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 7. Spatial Distribution of Infrared Land Surface Emissivity in the Taklimakan Desert for January Across Different Years and Wavelengths. (a) Emissivity at 8.3 μm in January 2001. (b) Emissivity at 8.3 μm in January 2023. (c) Difference in emissivity at 8.3 μm between January 2023 and January 2001. (d) Emissivity at 8.6 μm in January 2001. (e) Emissivity at 8.6 μm in January 2023. (f) Difference in emissivity at 8.6 μm between January 2023 and January 2001. (g) Emissivity at 9.1 μm in January 2001. (h) Emissivity at 9.1 μm in January 2023. (i) Difference in emissivity at 9.1 μm between January 2023 and January 2001. The color gradients represent land surface emissivity, with the scale adjusted to highlight the changes in emissivity over the two decades.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/d7d545596efbc3135c0a5635.png"},{"id":92931784,"identity":"db7b7896-08af-45f6-a985-8d7810235dd4","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":709985,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 8. Same with Figure 7 but for July.\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/dcfba576ac50260b3f1a7602.png"},{"id":92932159,"identity":"fe3e3b15-8689-4766-b73b-3e4f11e3a7b1","added_by":"auto","created_at":"2025-10-07 09:22:00","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":285093,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 9. Spatial Distribution and Temporal Changes in Soil Moisture for January and July from 2001 to 2023. (a) January 2001 soil moisture levels. (b) January 2023 soil moisture levels. (c) Difference in soil moisture between January 2023 and January 2001. (d) July 2001 soil moisture levels. (e) July 2023 soil moisture levels. (f) Difference in soil moisture between July 2023 and July 2001. The color gradients represent soil moisture in meters of water quivalent, with the scale adjusted to highlight changes over two decades.\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/26f3e3efacd64e821a6efc2b.png"},{"id":92931781,"identity":"9187ce92-24f7-42a9-adc2-1fc9c4705198","added_by":"auto","created_at":"2025-10-07 09:13:59","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":383143,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 10. Annual Trends in Statistical Metrics Reflecting the Independent Influence of Surface Temperature on Emissivity across Different Wavelengths from 2001 to 2022. (a) Partial correlation coefficients demonstrating the direct relationship between surface temperature and emissivity, controlling for soil moisture. (b) Multiple linear regression (MLR) coefficients for surface temperature, illustrating its individual impact on emissivity while accounting for soil moisture. (c) Random Forest (RF) importance scores for surface temperature, quantifying its predictive relevance to emissivity across wavelengths. (d) Interaction coefficients from MLR models, capturing the combined effect of surface temperature and soil moisture on emissivity. The distinct wavelengths are represented by different colors: blue for 8.3 μm, orange for 8.6 μm, and green for 9.1 μm.\u003c/p\u003e","description":"","filename":"image9.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/25a386d71e8a170a65b64c07.png"},{"id":92931794,"identity":"7c48406f-ee36-4db4-b759-0617396de0ed","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":302754,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 11: Spatial Distribution and Temporal Changes of SKT in the Study Area for January and July from 2001 to 2023. (a) January 2001 SKT. (b) January 2023 SKT. (c) The difference in SKT between January 2023 and January 2001. (d) July 2001 SKT. (e) July 2023 SKT. (f) The difference in SKT between July 2023 and July 2001. The color gradients represent surface temperature in Kelvin (K), with the scale adjusted to highlight the changes over the two decades.\u003c/p\u003e","description":"","filename":"image10.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/aa8dc81022ecf901bac8da8c.png"},{"id":92931791,"identity":"dec04149-7eb6-47b7-89f5-bec655438329","added_by":"auto","created_at":"2025-10-07 09:14:00","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":377119,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 12. Annual Variations in Statistical Metrics Reflecting the Independent Influence of Soil Moisture on Surface Emissivity across Different Wavelengths from 2001 to 2022. (a) Partial correlation coefficients indicating the direct relationship between soil moisture and emissivity, controlling for surface temperature. (b) Multiple linear regression (MLR) coefficients for soil moisture, showcasing its individual impact on emissivity while accounting for surface temperature. (c) Random Forest (RF) importance scores for soil moisture, quantifying its predictive relevance to emissivity across wavelengths. (d) Interaction coefficients from MLR models, capturing the combined effect of soil moisture and surface temperature on emissivity. The different wavelengths are represented by distinct colors: blue for 8.3 μm, orange for 8.6 μm, and green for 9.1 μm.\u003c/p\u003e","description":"","filename":"image11.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/8442372d32153ded128e1a39.png"},{"id":92932162,"identity":"0bb24194-0058-4efa-8a42-911c354b0a4e","added_by":"auto","created_at":"2025-10-07 09:22:00","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":481271,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 13. Comparative Analysis of Surface Emissivity in Relation to Soil Moisture and Surface Temperature for July 2001 and July 2023. (a) Soil moisture versus emissivity for July 2001, with correlation coefficients (R) for wavelengths 8.3 μm, 8.6 μm, and 9.1 μm indicated. (b) Surface temperature versus emissivity for July 2001, with correlation coefficients (R) for wavelengths 8.3 μm, 8.6 μm, and 9.1 μm indicated. (c) Soil moisture versus emissivity for July 2023, with updated correlation coefficients (R) for wavelengths 8.3 μm, 8.6 μm, and 9.1 μm indicated. (d) Surface temperature versus emissivity for July 2023, with updated correlation coefficients (R) for wavelengths 8.3 μm, 8.6 μm, and 9.1 μm indicated. Different colors represent different wavelengths as shown in the legend.\u003c/p\u003e","description":"","filename":"image12.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/c9d85f5971d39895667eb52a.png"},{"id":92932164,"identity":"86648ab9-7d3a-4b7a-a8dd-d8238face852","added_by":"auto","created_at":"2025-10-07 09:22:00","extension":"png","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":605962,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 14. 3D Scatter Plots Illustrating the Relationship Between Surface Emissivity, Soil Moisture, and Surface Temperature at Different Wavelengths for July 2007 and July 2023. (a-c) Data for July 2007 at wavelengths 8.3 μm, 8.6 μm, and 9.1 μm, respectively. (d-f) Data for July 2023 at wavelengths 8.3 μm, 8.6 μm, and 9.1 μm, respectively. The color gradient represents the range of surface emissivity values from 0.02 to 0.98.\u003c/p\u003e","description":"","filename":"image13.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/06c2e4f90ca2134a79700811.png"},{"id":92931816,"identity":"b65df5fd-f6ef-4acb-8b72-66004f816d5f","added_by":"auto","created_at":"2025-10-07 09:14:01","extension":"png","order_by":14,"title":"Figure 14","display":"","copyAsset":false,"role":"figure","size":288601,"visible":true,"origin":"","legend":"\u003cp\u003eFigure 15. Temporal Variations in the Influence of Soil Moisture on Emissivity Across Different Wavelengths (a) Coefficient of Soil Moisture in Multiple Linear Regression, (b) Coefficient of Interaction Term in Multiple Linear Regression, and (c) Importance of Soil Moisture in Random Forest Models.\u003c/p\u003e","description":"","filename":"image14.png","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/b7dd6a958f0bcd76c948d91a.png"},{"id":100614787,"identity":"a23cc333-be85-43c0-8c13-d2bd6e099e09","added_by":"auto","created_at":"2026-01-19 17:25:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4638169,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7587474/v1/1e8a5efc-448a-4b24-a085-359ad2d83231.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Infrared Land Surface Emissivity Dynamics in the Taklimakan Desert : Spatiotemporal Patterns and Key Drivers","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eLand surface emissivity (LSE) plays a pivotal role in governing thermal radiation exchange in hyper-arid regions, yet its spatiotemporal controls remain incompletely understood under extreme dryness. The Taklimakan Desert, characterized by persistent aridity and active aeolian processes, serves as an ideal natural laboratory to examine thermal versus hydrological drivers of LSE beyond diurnal timescales. Recent advances in remote sensing technology have enabled high-resolution analysis of emissivity variations across multiple spatiotemporal scales (Jin and Liang, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Previous research indicates LSE sensitivity to surface characteristics including soil moisture, vegetation cover, and mineral composition, as well as atmospheric conditions such as water vapor content and aerosol loading (Li, 2013). In hyper-arid environments like the Taklimakan Desert, where surface heterogeneity is pronounced (Zhang, 2015), quantifying these variations is essential for refining climate models, improving land surface temperature retrievals, and enhancing environmental monitoring capabilities (Chen, 2019).\u003c/p\u003e\u003cp\u003eExisting studies have established the significant impact of soil moisture on LSE, particularly in desert ecosystems (Mira, 2020; Hulley, 2020). Laboratory and satellite-based analyses demonstrate that even minor fluctuations in soil moisture can induce detectable emissivity changes (Ma, 2023), with cascading effects on energy partitioning and surface heat fluxes (Ogawa, 2008). While vegetation is sparse in deserts, its presence\u0026mdash;such as in oases and riparian zones\u0026mdash;modifies LSE through higher emissivity signatures (Sobrino, 2012; French, 2012). Additionally, atmospheric variability, including seasonal water vapor and dust perturbations, complicates emissivity retrieval, necessitating advanced correction approaches (Tang, 2021; Li, 2020).\u003c/p\u003e\u003cp\u003eDespite these advances, critical knowledge gaps persist regarding spatiotemporal LSE patterns in hyper-arid regions. Most existing studies have been limited to short-term observations or narrow spectral ranges, leaving long-term trends and wavelength-dependent behaviors insufficiently characterized (Wang, 2020; Zhou, 2022). Moreover, the synergistic effects of soil moisture, surface temperature, and land cover changes on LSE remain poorly quantified, particularly under climate change and anthropogenic pressures.\u003c/p\u003e\u003cp\u003eThe principal objectives of this study are to: (1) Analyze seasonal and interannual LSE variations at three infrared wavelengths; (2) Examine spatial distribution patterns and their relationships with land cover and climatic zones; (3) Quantify individual and combined influences of soil moisture and surface temperature using advanced statistical and machine learning techniques; and (4) Evaluate implications for regional climate modeling and remote sensing applications. By addressing these objectives, this research advances understanding of LSE dynamics in hyper-arid environments, supporting more accurate climate projections and sustainable land management strategies.\u003c/p\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1. Data Sources\u003c/h2\u003e\u003cp\u003eThis study utilized multiple datasets to analyze the spatiotemporal patterns of Infrared Land Surface Emissivity (LSE) in the Taklimakan Desert. The primary data sources include CAMEL Earth System Data Record (ESDR) and Ancillary Datasets (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://e4ftl01.cr.usgs.gov/MEASURES/CAM5K30EM.003/\u003c/span\u003e\u003cspan address=\"https://e4ftl01.cr.usgs.gov/MEASURES/CAM5K30EM.003/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe CAMEL Earth System Data Record (ESDR) is a global monthly land surface emissivity (LSE) database with a 5 km spatial resolution covering the spectral range of 3.6\u0026ndash;14.3 \u0026micro;m at 13 hinge points, which integrates the University of Wisconsin-Madison MODIS infrared emissivity dataset (UW BF) and the Jet Propulsion Laboratory ASTER Global Emissivity Dataset Version 4 (GEDv4) (Borbas, 2018; Wan, 2021). For this study, we specifically extracted LSE data at three key atmospheric window wavelengths (8.3 \u0026micro;m, 8.6 \u0026micro;m, and 9.1 \u0026micro;m) to analyze seasonal and interannual variations from 2001 to 2023.\u003c/p\u003e\u003cp\u003eTo complement the primary LSE analysis, we incorporated several ancillary datasets to examine key environmental drivers: (1) Soil moisture data were derived from the ERA5 monthly averaged Skin Reservoir Content (SRC), which provides high-resolution (0.25\u0026deg;\u0026times;0.25\u0026deg;) estimates of surface water storage in the top 7 cm soil layer, enabling precise assessment of hydrological influences on emissivity patterns (Hersbach, 2020; McColl, 2022). (2) Surface temperature data were obtained from the ERA5 monthly averaged Skin Temperature (SKT) product, a physically consistent reanalysis dataset that captures the radiative temperature of the land surface at 0.25\u0026deg; spatial resolution, allowing for robust quantification of thermal effects on LSE variations (Hersbach, 2023; Good, 2022). These ERA5-based datasets, produced by the European Centre for Medium-Range Weather Forecasts (ECMWF), integrate multiple satellite observations and in-situ measurements through advanced data assimilation techniques, ensuring high reliability for climate studies in arid regions like the Taklimakan Desert.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2. Methodology.\u003c/h2\u003e\u003cp\u003eThis study adopted a multi-scale analytical framework to systematically investigate the spatiotemporal patterns of Infrared Land Surface Emissivity (LSE) in the Taklimakan Desert (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) from 2001 to 2023. The methodology integrated statistical analysis (Mann, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1945\u003c/span\u003e; Kendall, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e1975\u003c/span\u003e), machine learning, and geospatial techniques to achieve research objectives through three-dimensional analysis.\u003c/p\u003e\u003cp\u003eFor temporal trend analysis, the study examined seasonal and interannual variations by calculating monthly LSE anomalies relative to long-term (2001\u0026ndash;2023) monthly means across three spectral bands (8.3\u0026micro;m, 8.6\u0026micro;m, and 9.1\u0026micro;m). Seasonal trends were quantified using linear regression and Theil-Sen estimators to minimize outlier effects, while key metrics including mean, median, interquartile range (IQR), and standard deviation were computed for each season (winter: December-February; summer: June-August). Spectral differences were identified through ANOVA and post-hoc Tukey tests (Wang, 2023).\u003c/p\u003e\u003cp\u003eSpatial pattern characterization employed pixel-based analysis. The Mann-Kendall test identified significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) monotonic LSE trends for each 5km grid cell, with Sen's slope estimating change magnitude (Sen, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e1968\u003c/span\u003e). Getis-Ord\u0026rsquo;s statistics delineated high-change hotspots (ΔLSE\u0026thinsp;\u0026gt;\u0026thinsp;0.05 during 2001\u0026ndash;2023) (Getis, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e1992\u003c/span\u003e). Furthermore, LSE trends were stratified by MODIS land cover types (desert, oasis) to assess surface-type dependencies (Wang, 2023).\u003c/p\u003e\u003cp\u003eFor factor quantification, the study focused on soil moisture (SRC) and surface temperature (SKT) impacts. Partial correlation analysis isolated independent SRC and SKT effects while controlling for elevation and other covariates. Machine learning approaches, including Random Forest (1000 trees, 10-fold cross-validation) and Multiple Linear Regression, were applied to evaluate variable importance, capture nonlinear relationships, and quantify coefficients (including SRC\u0026times;SKT interaction terms). Sensitivity scenarios simulated LSE responses under controlled conditions: SRC-only changes with fixed SKT (2001\u0026ndash;2023 mean) and SKT-only changes with fixed SRC (hyper-arid baseline: 1\u0026times;10⁻⁶), providing insights into environmental factors' relative contributions to emissivity variations.\u003c/p\u003e\u003cp\u003eTo quantify the nonlinear relationships between environmental variables (soil moisture, surface temperature) and land surface emissivity (LSE), this study employed Random Forest (RF) regression and Multiple Linear Regression (MLR) with interaction terms. The RF algorithm, implemented via Python\u0026rsquo;s scikit-learn library, utilized 1,000 decision trees to ensure robust feature importance estimation and generalization. Hyperparameter optimization included limiting tree depth to 15 nodes and setting a minimum leaf sample size of 5 to prevent overfitting. Ten-fold cross-validation was applied to partition the dataset (2001\u0026ndash;2023, N\u0026thinsp;=\u0026thinsp;3,625 annual grid-cell observations), ensuring model performance evaluation across spatially independent subsets. Feature importance scores were derived using permutation-based methods, quantifying the relative contribution of soil moisture (SRC), surface temperature (SKT), and their interactions to LSE variability. For MLR, interaction terms (SRC \u0026times; SKT) were incorporated to capture synergistic effects, with coefficients standardized using z-score normalization to address multicollinearity. Sensitivity analyses further isolated individual factor contributions by fixing covariates (e.g., simulating SRC-only changes under mean SKT conditions). This hybrid approach leverages machine learning\u0026rsquo;s capacity to detect nonlinear patterns while maintaining interpretability through traditional regression frameworks, ensuring comprehensive quantification of LSE drivers in hyper-arid environments.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e3.1. Monthly and Interannual Variation.\u003c/h2\u003e\u003cp\u003eThis section examines monthly and interannual LSE patterns based solely on satellite-derived data (2001\u0026ndash;2023), excluding short-term diurnal effects (Ma, 2025).\u003c/p\u003e\u003cdiv id=\"Sec7\" class=\"Section3\"\u003e\u003ch2\u003e3.1.1 Monthly variation\u003c/h2\u003e\u003cp\u003eThe monthly analysis presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e3\u003c/span\u003e further elucidates these temporal patterns through a detailed examination of intra-annual cycles. At 8.3 \u0026micro;m wavelength, the average emissivity exhibits pronounced seasonal variations, fluctuating between 0.82 (winter minimum) and 0.94 (summer maximum), with no significant long-term trend detected over the 23-year period. The 8.6 \u0026micro;m wavelength mirrors this seasonal pattern but with slightly reduced amplitude (0.84\u0026ndash;0.92), suggesting wavelength-dependent sensitivity to surface thermal processes. Notably, the 9.1 \u0026micro;m data demonstrate the most constrained seasonal variation (0.84\u0026ndash;0.90), supporting the hypothesis that longer wavelengths are less responsive to seasonal surface changes in hyper-arid environments.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe monthly resolution analysis in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e4\u003c/span\u003e provides additional granularity to these temporal patterns. Winter months (December-February) consistently show elevated median emissivity values across all wavelengths (e.g., 0.883 at 8.3 \u0026micro;m in December), likely due to reduced thermal contrast between surface components. Summer months (June-August) exhibit peak values (e.g., 0.893 at 8.3 \u0026micro;m in July) that correlate with maximum surface temperatures. Transitional seasons display intermediate values, with spring (March-May) typically showing slightly lower emissivity than autumn (September-November), possibly reflecting differences in surface preconditioning from winter cooling versus summer heating.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section3\"\u003e\u003ch2\u003e3.1.2 Interannual variation\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e5\u003c/span\u003e presents a comprehensive analysis of emissivity measurements across three atmospheric window wavelengths (8.3, 8.6, and 9.1 \u0026micro;m) from 2001 to 2023, revealing distinct seasonal, spectral, and interannual patterns that reflect the complex thermal dynamics of the Taklimakan Desert surface. The observed emissivity exhibited consistent seasonal variability across all wavelengths, with summer displaying the highest values (e.g., ~\u0026thinsp;0.89 at 8.3 \u0026micro;m) and winter the lowest (~\u0026thinsp;0.87 at 8.3 \u0026micro;m). This pattern aligns with theoretical expectations for arid surfaces, where increased summer temperatures enhance thermal emission while winter cooling reduces radiative efficiency. Intermediate values during spring and autumn transition periods suggest gradual surface property changes in response to seasonal climate shifts.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eBuilding upon these seasonal trends, spectral dependence emerged as a key characteristic of the dataset. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e5\u003c/span\u003e, emissivity systematically decreased with increasing wavelength, with 8.3 \u0026micro;m consistently recording the highest averages (~\u0026thinsp;0.87), followed by 8.6 \u0026micro;m (~\u0026thinsp;0.86) and 9.1 \u0026micro;m (~\u0026thinsp;0.84). This spectral gradient corresponds to known material-specific emissivity properties, where shorter wavelengths typically yield higher emissivity for silicate-dominated desert surfaces due to reduced atmospheric absorption and stronger surface-atmosphere coupling.\u003c/p\u003e\u003cp\u003eTransitioning to interannual variability, the three wavelengths exhibited divergent long-term behaviors that warrant detailed examination. At 8.3 \u0026micro;m, a gradual but statistically significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) increase in emissivity was observed from 2001 to 2023, with a mean decadal trend of +\u0026thinsp;0.008\u0026thinsp;\u0026plusmn;\u0026thinsp;0.002. In contrast, 8.6 \u0026micro;m displayed greater interannual variability, characterized by sharp peaks in the early 2000s followed by an upward trend in summer and autumn months post-2010. The 9.1 \u0026micro;m data showed relative stability with minor fluctuations, suggesting this wavelength may be less sensitive to surface changes over the study period.\u003c/p\u003e\u003cp\u003eNotable anomalies in the time series provide additional insights into extreme events and their impacts on surface properties. The emissivity spikes observed in January 2005 and 2008 coincide with the documented extreme cold waves and associated surface disturbances in the Taklimakan Desert. During January 2005, the record-breaking low temperatures (e.g., \u0026minus;\u0026thinsp;28.6\u0026deg;C at Ruoqiang Station) likely induced thermal contraction of surface minerals, particularly quartz-dominated sands, which altered micro-scale roughness and cavity radiation efficiency. This process temporarily elevated emissivity at shorter wavelengths (e.g., 8.3 \u0026micro;m), where sensitivity to surface thermal changes is highest, resulting in localized LSE increases of up to 0.04 compared to adjacent months. Concurrently, sporadic dust events in December 2005\u0026mdash;triggered by dry conditions and strong winds\u0026mdash;deposited fine particles that modified surface spectral properties, further amplifying emissivity anomalies in central dune fields. In January 2008, despite the absence of direct snowfall, prolonged cold anomalies (e.g., \u0026minus;\u0026thinsp;22.1\u0026deg;C at Hotan Station) intensified radiative cooling, reducing surface thermal inertia. This led to sharper diurnal temperature gradients, which enhanced thermal stress on surface materials. The resulting microfractures and particle disaggregation increased cavity radiation, particularly at 9.1 \u0026micro;m, where emissivity rose by 0.03\u0026ndash;0.05. Additionally, residual dust from autumn 2007 (due to reduced precipitation) formed a thin, transient layer that altered shortwave absorption and longwave emission characteristics. These combined effects explain the anomalous emissivity peaks during both winters, demonstrating that surface temperature extremes, even under stable soil moisture conditions, can drive significant LSE variability through thermal-mechanical and aeolian processes. Critically, while soil moisture remains a foundational control on emissivity in arid systems, these cases highlight that temperature-driven surface changes can dominate LSE anomalies during extreme cold events. This underscores the need to integrate thermal stress models into emissivity retrieval algorithms, especially for regions prone to rapid temperature fluctuations.\u003c/p\u003e\u003cp\u003eTo quantify these temporal variations more precisely, Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e6\u003c/span\u003e presents boxplot analyses of annual emissivity distributions. The 8.3 \u0026micro;m wavelength shows the broadest range of variation (0.86\u0026ndash;0.92), with a notable peak in 2023 that may reflect exceptional surface conditions or climatic anomalies. The 8.6 \u0026micro;m data exhibit similar interannual patterns but with reduced magnitude (0.85\u0026ndash;0.88), while 9.1 \u0026micro;m displays the most stable annual distributions (0.84\u0026ndash;0.87). This wavelength hierarchy persists throughout the study period, confirming the spectral dependence of emissivity variability in desert environments.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThese temporal analyses collectively demonstrate that LSE variations in the Taklimakan Desert are governed by three principal factors: (1) seasonal thermal cycles that modulate surface emission properties, (2) wavelength-dependent sensitivity to surface characteristics, and (3) long-term environmental changes that induce gradual emissivity trends. The consistency of these patterns across multiple analytical approaches (time series, boxplots, and monthly distributions) reinforces the robustness of these findings and their relevance for understanding land-atmosphere interactions in arid regions.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e3.2. Spatial Distribution Characteristics\u003c/h2\u003e\u003cp\u003eBuilding upon the temporal analysis, this section systematically examines the spatial heterogeneity and decadal evolution of infrared land surface emissivity (LSE) across the Taklimakan Desert. A comprehensive quantitative analysis was conducted for three atmospheric window wavelengths (8.3 \u0026micro;m, 8.6 \u0026micro;m, and 9.1 \u0026micro;m) using observations from July 2001 and 2023. Spatial difference maps were generated to assess decadal trends (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e7\u003c/span\u003e), revealing wavelength-specific responses to surface processes.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eAt 8.3 \u0026micro;m, January 2001 exhibited a mean emissivity of 0.890 (median: 0.880, range: 0.835\u0026ndash;0.987), with interquartile values (Q1\u0026ndash;Q3) spanning 0.865\u0026ndash;0.905. By 2023, significant increases were observed, particularly in central and western regions (Δ\u0026thinsp;=\u0026thinsp;0.04\u0026ndash;0.08). Approximately 72% of the desert area showed emissivity increases exceeding 0.05, with the most pronounced changes (Δ\u0026thinsp;\u0026gt;\u0026thinsp;0.07) concentrated in active dune fields. This spatial clustering suggests aeolian processes may dominate emissivity changes at shorter wavelengths.\u003c/p\u003e\u003cp\u003eTransitioning to 8.6 \u0026micro;m, baseline measurements for January 2001 showed slightly lower emissivity (mean: 0.881, median: 0.867, range: 0.832\u0026ndash;0.981) compared to 8.3 \u0026micro;m. The difference analysis revealed heterogeneous spatial responses: central regions exhibited increases up to 0.07, while western margins demonstrated decreases of 0.02\u0026ndash;0.03. This contrast highlights the wavelength's sensitivity to surface mineralogical variations, potentially influenced by differential quartz and feldspar distributions across geomorphic units.\u003c/p\u003e\u003cp\u003eThe 9.1 \u0026micro;m data displayed the lowest baseline emissivity (mean: 0.868, median: 0.851, range: 0.816\u0026ndash;0.983), consistent with spectral absorption characteristics of silicate minerals. Difference analysis identified the strongest warming signals in central regions (Δ\u0026thinsp;\u0026gt;\u0026thinsp;0.08), while eastern margins showed minimal change (Δ\u0026thinsp;\u0026lt;\u0026thinsp;0.02). Notably, the interquartile range narrowed from 0.025 in 2001 to 0.018 in 2023, indicating a trend toward surface property homogenization at longer wavelengths.\u003c/p\u003e\u003cp\u003eThe multi-wavelength comparison revealed consistent emissivity increases across all bands from Jan 2001 to Jan 2023, with mean differences (Δ) of +\u0026thinsp;0.062 at 8.3 \u0026micro;m, +\u0026thinsp;0.058 at 8.6 \u0026micro;m, and +\u0026thinsp;0.053 at 9.1 \u0026micro;m. Spatial variability decreased markedly, as evidenced by 18\u0026ndash;22% reductions in standard deviations across wavelengths. Regional hotspots with Δ\u0026thinsp;\u0026gt;\u0026thinsp;0.07 were predominantly located in central dune fields, correlating with areas of intense aeolian activity documented in field surveys.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo contextualize these spatial patterns, Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e8\u003c/span\u003e provides a detailed quantitative assessment of July emissivity distributions across the study period. At 8.3 \u0026micro;m (Panels a\u0026ndash;c), the 2023 data showed a 7.8% expansion of high-emissivity zones (\u0026gt;\u0026thinsp;0.90) compared to 2001, particularly along southern margin oases. The 8.6 \u0026micro;m difference map (Panel f) revealed a dipole pattern, with central increases (+\u0026thinsp;0.07) offset by western decreases (-0.03), suggesting competing influences of surface darkening and mineralogical sorting. At 9.1 \u0026micro;m (Panels g\u0026ndash;i), the observed Δ\u0026thinsp;\u0026gt;\u0026thinsp;0.08 in western regions coincided with areas of reduced albedo (-9% based on MODIS data), emphasizing the wavelength's sensitivity to surface radiative properties.\u003c/p\u003e\u003cp\u003eThe potential drivers and broader significance of the observed trends can be attributed to multiple factors. Surface warming is a key contributor, as regional temperature records indicate a decadal increase of approximately\u0026thinsp;+\u0026thinsp;1.2\u0026deg;C. Furthermore, changes in surface roughness due to aeolian processes may alter cavity radiation properties, thereby influencing emissivity. These factors collectively contribute to the observed changes in emissivity and surface properties, highlighting the complex interplay between climatic and geomorphological processes in shaping the region's thermal characteristics.\u003c/p\u003e\u003cp\u003eIt is worth noting that similar spatial trends were observed in April and October, with the central and southern regions experiencing emissivity increases comparable to those in July. However, the magnitude of these changes was smaller, indicating that seasonal variations may influence the extent of emissivity alterations. This suggests that while the general pattern of increasing emissivity persists throughout these months, the intensity of the changes is most pronounced in July, likely due to peak thermal conditions during this period. These findings underscore the utility of multi-wavelength LSE analysis in monitoring arid land surface dynamics and their responses to climate change. The consistent emissivity increases suggest a general warming and drying trend, with implications for regional climate modeling, energy balance studies, and remote sensing applications.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Analysis of Influencing Factors\u003c/h2\u003e\u003cdiv id=\"Sec11\" class=\"Section3\"\u003e\u003ch2\u003e3.3.1 The independent effect of soil moisture on emissivity.\u003c/h2\u003e\u003cp\u003eTo isolate the hydrological influence on emissivity, we first analyzed the spatiotemporal evolution of soil moisture content (SRC) across 3,625 grid points in the Taklimakan Desert (2001\u0026ndash;2023). As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e9\u003c/span\u003e, SRC values exhibited extreme aridity (order of 10⁻⁶) with pronounced seasonal contrasts: Winter (January) means were 3.6 times higher than summer (July) values (1.45\u0026times;10⁻⁶ vs. 4.04\u0026times;10⁻⁷ m of water quivalent in 2001). However, a significant desiccation trend emerged over the 22-year period, with January SRC declining by 61.5% (5.57\u0026times;10⁻⁷ in 2023) compared to a modest summer reduction of 1.65\u0026times;10⁻⁷.\u003c/p\u003e\u003cp\u003eSpatially, winter drying exhibited broad-scale uniformity (std reduction: 1.14\u0026times;10⁻⁶ \u0026rarr; 3.63\u0026times;10⁻⁷), while summer changes showed localized extremes (range: -1.60\u0026times;10⁻⁵ to +\u0026thinsp;1.17\u0026times;10⁻⁵). The interquartile range (IQR) narrowed by 78% in winter versus a negatively skewed summer distribution, indicating distinct moisture loss mechanisms: Winter declines likely reflect weakened westerly moisture transport under Arctic Oscillation modulation, whereas summer variability stems from oasis irrigation and sporadic rainfall.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eBuilding upon these spatial patterns, soil moisture-emissivity relationships were quantified through partial correlation analysis controlling for temperature. The persistent positive correlations (R\u0026thinsp;=\u0026thinsp;0.475\u0026ndash;0.527 in 2001) weakened by 2023 (R\u0026thinsp;=\u0026thinsp;0.315\u0026ndash;0.378), suggesting diminishing hydrological control under intensified aridity. Machine learning models revealed that a 1-unit SRC increase elevated emissivity by 0.028\u0026thinsp;\u0026plusmn;\u0026thinsp;0.018 (peak sensitivity at 9.1 \u0026micro;m: 0.035\u0026thinsp;\u0026plusmn;\u0026thinsp;0.020), though post-2015 coefficients declined by ~\u0026thinsp;40%, indicative of ecosystem threshold behavior (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e10\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eTo contextualize these hydrological changes, seasonal divergence was critically examined. Winter\u0026rsquo;s stronger emissivity coupling (55\u0026ndash;72% RF importance) aligns with moisture retention in colder soils, whereas summer\u0026rsquo;s attenuated effects correlate with rapid evaporation under extreme heat. The narrowing IQR in winter SRC (-1.33\u0026times;10⁻⁶ to -1.64\u0026times;10⁻⁷) versus summer\u0026rsquo;s localized extremes (-1.60\u0026times;10⁻⁵ to +\u0026thinsp;1.17\u0026times;10⁻⁵) further confirms that large-scale atmospheric drivers dominate cold-season drying, while warm-season variations are governed by microscale land-atmosphere interactions.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThese findings provide critical insights into the \"winter-dominated, summer-attenuated\" drying paradigm, demonstrating that even hyper-arid systems exhibit nuanced moisture-emissivity coupling. The results emphasize the necessity of seasonally adaptive emissivity parameterizations in climate models, particularly for capturing winter\u0026rsquo;s stronger hydrological signals in energy balance calculations.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\u003ch2\u003e3.3.2 The independent effect of surface temperature on emissivity.\u003c/h2\u003e\u003cp\u003eTo further investigate the factors influencing land surface emissivity (LSE) in the Taklimakan Desert, this section examines the independent effect of surface temperature (SKT) on LSE. This analysis builds upon the previous section's findings on the influence of soil moisture, providing a comprehensive understanding of the individual contributions of these key environmental variables.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e11\u003c/span\u003e: Spatial Distribution and Temporal Changes of SKT in the Study Area for January and July from 2001 to 2023. (a) January 2001 SKT. (b) January 2023 SKT. (c) The difference in SKT between January 2023 and January 2001. (d) July 2001 SKT. (e) July 2023 SKT. (f) The difference in SKT between July 2023 and July 2001. The color gradients represent surface temperature in Kelvin (K), with the scale adjusted to highlight the changes over the two decades.\u003c/p\u003e\u003cp\u003eIn January (panels a-c), the mean SKT decreased from 267.09 K in 2001 to 264.91 K in 2023, a change of -2.18 K. The median decline was \u0026minus;\u0026thinsp;1.87 K, and spatial variability was reduced, as indicated by a standard deviation decrease from 1.52 K to 1.08 K. Extreme values showed localized cooling up to -7.10 K and limited warming up to +\u0026thinsp;2.85 K, supported by a downward-shifted interquartile range (IQR) with q25 at -2.46 K and q75 at -1.58 K. In contrast, July (panels d-f) exhibited warming, with the mean SKT increasing from 306.69 K in 2001 to 307.98 K in 2023, a change of +\u0026thinsp;1.28 K. The median increase was +\u0026thinsp;1.17 K, and spatial heterogeneity increased, as indicated by a standard deviation increase from 2.70 K to 3.23 K. Maximum warming reached\u0026thinsp;+\u0026thinsp;4.09 K, while minor cooling was observed at -2.33 K, consistent with a positive IQR shift (q25: +0.43 K, q75: +2.05 K).\u003c/p\u003e\u003cp\u003eThese divergent trends\u0026mdash;January cooling versus July warming\u0026mdash;highlight seasonally asymmetric responses to climatic or anthropogenic drivers. The reduced January variability may indicate homogenizing influences, such as changes in cloud cover, while the increased variability in July could reflect enhanced surface-atmosphere feedbacks. These findings align with global climate change narratives but emphasize localized heterogeneity. This quantitative analysis underscores the utility of long-term SKT monitoring to disentangle seasonal and spatial climate signals. Future work should integrate ancillary data, such as land cover and albedo, to attribute observed changes and refine predictive models. Such insights are critical for ecosystem management and climate adaptation strategies in thermally sensitive regions.\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e12\u003c/span\u003e provides a comprehensive quantitative analysis of soil moisture's influence on surface emissivity across three thermal infrared wavelengths (8.3 \u0026micro;m, 8.6 \u0026micro;m, and 9.1 \u0026micro;m) from 2001 to 2022, employing four complementary statistical approaches. Partial correlation coefficients (Panel a) reveal persistent positive relationships (ranging 0.14\u0026ndash;0.48 for 8.3 \u0026micro;m, 0.18\u0026ndash;0.54 for 8.6 \u0026micro;m, and 0.15\u0026ndash;0.52 for 9.1 \u0026micro;m) after controlling for surface temperature, with peak correlations occurring in 2014\u0026ndash;2018, particularly at longer wavelengths. The multiple linear regression (MLR) coefficients (Panel b) demonstrate wavelength-dependent sensitivity, where soil moisture's marginal effect on emissivity remained stable yet spectrally distinct (0.01\u0026ndash;0.07 for 8.3 \u0026micro;m, 0.01\u0026ndash;0.07 for 8.6 \u0026micro;m, 0.01\u0026ndash;0.07 for 9.1 \u0026micro;m), with notable stability during 2005\u0026ndash;2015. Random Forest importance scores (Panel c) highlight soil moisture's predictive dominance (55\u0026ndash;72% relative importance) at 8.3 \u0026micro;m compared to other wavelengths, though all bands showed interannual variability (\u0026plusmn;\u0026thinsp;15% importance). Interaction terms (Panel d) expose complex nonlinearities, with combined soil moisture-temperature effects strongest at 9.1 \u0026micro;m (coefficients reaching \u0026minus;\u0026thinsp;6.25 in 2019) and weakest at 8.3 \u0026micro;m (-3.24 in 2023). Spectral differences emerge clearly: 8.6 \u0026micro;m consistently showed the strongest partial correlations (e.g., 0.54 in 2014), while 9.1 \u0026micro;m exhibited the most pronounced interaction effects, suggesting wavelength-specific coupling between hydrological and thermal processes. These findings, derived from 24 years of continuous data (N\u0026thinsp;=\u0026thinsp;3,625 annual observations per wavelength), underscore soil moisture's role as a key modulator of emissivity, with implications for refining land-atmosphere models and hyperspectral remote sensing algorithms, particularly in addressing wavelength-dependent biases in current emissivity parameterizations.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section3\"\u003e\u003ch2\u003e3.3.3 The joint effect of soil moisture and surface temperature on emissivity.\u003c/h2\u003e\u003cp\u003eTo provide a comprehensive understanding of the factors influencing land surface emissivity (LSE) in the Taklimakan Desert, this section examines the combined effects of soil moisture (SRC) and surface temperature (SKT) on LSE. This analysis builds upon the previous sections' findings on the individual contributions of these variables, offering insights into their synergistic impacts.\u003c/p\u003e\u003cp\u003eThe relationship between surface emissivity and both soil moisture and surface temperature was analyzed across two distinct periods, July 2001 and July 2023. The results reveal significant insights into the emissivity characteristics at different wavelengths (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e13\u003c/span\u003e). The study demonstrates that soil moisture is positively correlated with surface emissivity across the 8.3 \u0026micro;m, 8.6 \u0026micro;m, and 9.1 \u0026micro;m wavelengths for both years. In July 2001, the correlation coefficients were 0.475, 0.510, and 0.527, respectively, while in July 2023, they were slightly lower at 0.315, 0.367, and 0.378. Despite the maintained positive correlation, the decrease in these coefficients over time may suggest an influence of climate change or alterations in land use patterns.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFurther linear regression analysis elucidates the impact of soil moisture on surface emissivity. The slopes of the regression lines for July 2001 were 0.067, 0.067, and 0.075, indicating a stronger influence of soil moisture on emissivity compared to July 2023, where the slopes were 0.041, 0.044, and 0.050. These results, which are statistically significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), suggest that the effect of soil moisture on emissivity has diminished over the two decades.\u003c/p\u003e\u003cp\u003eConversely, the relationship between surface temperature and surface emissivity exhibits a notable negative correlation across all three wavelengths for both years. The correlation coefficients for July 2001 were \u0026minus;\u0026thinsp;0.493, -0.544, and \u0026minus;\u0026thinsp;0.565, while for July 2023 they were \u0026minus;\u0026thinsp;0.442, -0.504, and \u0026minus;\u0026thinsp;0.517. Although the negative correlation remains robust, a slight reduction in these coefficients from 2001 to 2023 is observed.\u003c/p\u003e\u003cp\u003eLinear regression analysis also indicates that the negative influence of surface temperature on surface emissivity is significant (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) for both periods. The slopes for July 2001 were \u0026minus;\u0026thinsp;0.004, -0.004, and \u0026minus;\u0026thinsp;0.005, and for July 2023, they were \u0026minus;\u0026thinsp;0.003, -0.003, and \u0026minus;\u0026thinsp;0.003, respectively. The slight decrease in these slopes over time suggests a reduced negative impact of surface temperature on emissivity.\u003c/p\u003e\u003cp\u003eIn conclusion, Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e11\u003c/span\u003e highlights changes in the relationships between surface emissivity and both soil moisture and surface temperature over a 22-year period. The observed variations in correlation and regression slopes may reflect the effects of climate change and land use alterations on surface emissivity. Future research should further investigate the driving factors behind these changes and their potential impacts on the surface energy balance.\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e14\u003c/span\u003e presents a comparative analysis of the relationships among Land Surface Emissivity (LSE), Soil Moisture (SRC), and Surface Temperature (SKT) at three distinct wavelengths (8.3 \u0026micro;m, 8.6 \u0026micro;m, and 9.1 \u0026micro;m) for the years 200107 and 202307. Each subplot corresponds to a specific wavelength, with subplots (a), (b), and (c) depicting data from the year 200107, and subplots (d), (e), and (f) illustrating data from the year 202307. The three-dimensional scatter plots provide a visual representation of the interdependencies among these variables and their temporal evolution.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eA discernible pattern in the distribution of LSE with varying wavelengths is evident from Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e14\u003c/span\u003e. At wavelengths of 8.3 \u0026micro;m and 8.6 \u0026micro;m, the relationship between LSE and both SRC and SKT appears to be more pronounced, whereas at 9.1 \u0026micro;m, the relationship seems to diminish. This variation may be attributed to the differential sensitivity of these wavelengths to surface characteristics. Furthermore, a comparison of data from 200107 and 202307 reveals discrepancies in the range and central tendency of LSE at the same wavelengths, potentially reflecting changes in surface properties over the intervening period.\u003c/p\u003e\u003cp\u003eUpon closer examination of the scatter distributions in Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e14\u003c/span\u003e, it is observed that regions of higher LSE consistently correspond to lower SRC and higher SKT across all wavelengths. This phenomenon may be associated with the direct impact of SRC on LSE, where lower SRC values coincide with higher LSE values. Additionally, elevated SKT may contribute to increased LSE. These findings are consistent with existing literature, further substantiating the intricate interplay between LSE, SRC, and SKT.\u003c/p\u003e\u003cp\u003eThe insights gleaned from Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e14\u003c/span\u003e underscore the importance of considering wavelength-specific interactions when analyzing surface properties. The observed changes in LSE over time may have significant implications for remote sensing applications and understanding surface processes. Future research endeavors could delve deeper into the causal mechanisms underlying these relationships and explore additional factors that might influence LSE, such as vegetation cover and land use patterns.\u003c/p\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig14\" class=\"InternalRef\"\u003e15\u003c/span\u003e illustrates the temporal variations in the influence of soil moisture on land surface emissivity across three spectral wavelengths (8.3 \u0026micro;m, 8.6 \u0026micro;m, and 9.1 \u0026micro;m) derived from multiple analytical approaches. Panel (a) presents the soil moisture coefficients from multiple linear regression (MLR) models, demonstrating distinct annual fluctuations. At 8.3 \u0026micro;m, coefficients ranged from 0.0027 (2011) to 0.0745 (2003), with notable interannual variability linked to climatic events. Across all wavelengths (2001\u0026ndash;2023), soil moisture increases raised emissivity by 0.028\u0026thinsp;\u0026plusmn;\u0026thinsp;0.018 units per 1-unit moisture increase, with the strongest sensitivity observed at 9.1 \u0026micro;m (average: 0.035\u0026thinsp;\u0026plusmn;\u0026thinsp;0.020 units, peaking at 0.074 units in 2003). Post-2015, coefficients declined by ~\u0026thinsp;40% (e.g., 9.1 \u0026micro;m: 0.035 to 0.021), reflecting reduced soil moisture-emissivity coupling under aridification.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe interaction term coefficients between soil moisture and temperature (Panel b) revealed complex non-linear relationships. Negative values dominated across wavelengths (e.g., -5.87 in 2019 at 8.3 \u0026micro;m), indicating antagonistic effects between soil moisture and thermal conditions on emissivity. Temperature increases reduced emissivity by -0.0029\u0026thinsp;\u0026plusmn;\u0026thinsp;0.0012 units per 1\u0026deg;C rise, with the strongest cooling at 9.1 \u0026micro;m (-0.0035\u0026thinsp;\u0026plusmn;\u0026thinsp;0.0015 units). Interaction terms partially offset temperature-driven reductions by ~\u0026thinsp;15% (e.g., net effect at 9.1 \u0026micro;m: -0.0030 vs. raw \u0026minus;\u0026thinsp;0.0035). The 9.1 \u0026micro;m band displayed amplified interaction magnitudes (e.g., -6.25 in 2019), aligning with stronger water absorption at longer infrared wavelengths. Temporal patterns exhibited episodic spikes (e.g., 2003, 2019) corresponding to extreme drought events.\u003c/p\u003e\u003cp\u003eRandom Forest feature importance metrics (Panel c) corroborated the MLR findings, with soil moisture accounting for 60.9\u0026ndash;72.6% of explanatory power across wavelengths. The 8.6 \u0026micro;m band showed the most stable importance values (\u0026plusmn;\u0026thinsp;5.8% interannual variation), while 9.1 \u0026micro;m demonstrated the strongest correlation with soil moisture anomalies (R\u0026sup2;=0.83 against satellite products). Divergences between MLR and Random Forest results (e.g., 12\u0026ndash;18% larger soil moisture impacts in non-linear models) highlight machine learning\u0026rsquo;s advantage in capturing complex interactions, particularly post-2015. Collectively, these multi-method analyses quantify soil moisture\u0026rsquo;s dominant role (emissivity increase: ~0.035 units per 1-unit moisture at 9.1 \u0026micro;m) and temperature\u0026rsquo;s secondary cooling effect (-0.0035 units per 1\u0026deg;C), emphasizing the critical role of wavelength selection in thermal remote sensing of land surface processes.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThe spatiotemporal dynamics of land surface emissivity (LSE) in the hyper-arid Taklimakan Desert underscore the complex interplay between soil moisture (SRC), surface temperature (SKT), and surface radiative properties. While prior studies have established soil moisture as a primary driver of LSE variability in arid regions (Hulle, 2020; Mira, 2020), our findings highlight a critical nuance: even under negligible soil moisture changes, LSE exhibits significant fluctuations driven by surface temperature variations. This does not contradict the established role of soil moisture but rather expands the understanding of emissivity controls in extreme aridity, where thermal processes dominate due to suppressed hydrological activity.\u003c/p\u003e\u003cp\u003eConsistent with existing literature, soil moisture remains a foundational factor in modulating LSE. Our results confirm its positive correlation with emissivity (e.g., R\u0026thinsp;=\u0026thinsp;0.475\u0026ndash;0.527 in 2001), particularly at longer wavelengths (9.1 \u0026micro;m: 0.035\u0026thinsp;\u0026plusmn;\u0026thinsp;0.020 LSE increase per unit SRC). However, in the hyper-arid Taklimakan Desert, where soil moisture is chronically low (\u0026le;\u0026thinsp;2 g/kg) and exhibits minimal interannual variability (ΔSRC\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;0.15 g/kg decade⁻\u0026sup1;), the hydrological influence is constrained. This aligns with McColl (2022), who noted nonlinear moisture-emissivity coupling in water-limited systems, where soil moisture effects saturate under extreme aridity.\u003c/p\u003e\u003cp\u003eThe observed decadal LSE increase (+\u0026thinsp;0.053\u0026ndash;+0.062 decade⁻\u0026sup1;) despite regional drying and warming highlights the dominance of aeolian and thermal processes over hydrological drivers. While soil moisture decline would theoretically reduce emissivity, SKT-driven reductions (\u0026minus;\u0026thinsp;0.0029\u0026thinsp;\u0026plusmn;\u0026thinsp;0.0012 per 1\u0026deg;C) are counteracted by surface darkening (Δalbedo\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;7%) and particle sorting in active dunes. Central hotspots (ΔLSE\u0026thinsp;\u0026gt;\u0026thinsp;0.07) correlate with enhanced cavity radiation from wind-driven mineral redistribution, a feedback mechanism amplified by warming (Zhou, 2022). This synergy between thermal and geomorphological processes underscores the need to expand emissivity models beyond soil moisture-centric frameworks in hyper-arid environments.\u003c/p\u003e\u003cp\u003eThe wavelength-dependent characteristics of land surface emissivity (LSE) offer critical insights for both remote sensing applications and climate modeling. The interannual stability of the 8.6 \u0026micro;m band (CV\u0026thinsp;=\u0026thinsp;1.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.3%) makes it particularly suited for establishing baseline emissivity products. In contrast, the 9.1 \u0026micro;m band exhibits the strongest interaction between surface temperature and soil moisture (MLR coefficient\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;6.25), rendering it invaluable for detecting combined thermal-hydrological stresses. Seasonally, the enhanced coupling of soil moisture and emissivity in winter (RF importance\u0026thinsp;=\u0026thinsp;55\u0026ndash;72%) starkly contrasts with summer\u0026rsquo;s dominance of thermal-driven LSE variations, underscoring the necessity for climate models to adopt seasonally adaptive algorithms.\u003c/p\u003e\u003cp\u003eThese findings refine the contextual relevance of soil moisture in hyper-arid regions, where temperature and aeolian processes supersede hydrological controls. For missions like ESA\u0026rsquo;s LSTM, incorporating wavelength-specific SKT corrections (e.g., \u0026minus;\u0026thinsp;0.0035\u0026thinsp;\u0026plusmn;\u0026thinsp;0.0015 LSE per 1\u0026deg;C at 9.1 \u0026micro;m) is critical to minimize retrieval biases. Future work should integrate high-resolution hyperspectral data (e.g., EMIT) to resolve microscale heterogeneity.\u003c/p\u003e\u003cp\u003eThis study reaffirms soil moisture as a key LSE driver in arid systems while demonstrating that surface temperature variations can independently modulate emissivity under stable hydrological conditions. By quantifying these dual controls, we advance emissivity parameterizations for climate models and remote sensing, particularly in addressing the unique thermal-geomorphological feedbacks of hyper-arid deserts.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eKey findings reveal that even under minimal soil moisture variations, surface temperature fluctuations significantly influence LSE, highlighting their importance in arid environments.\u003c/p\u003e\u003cp\u003eThe spectral characteristics of LSE underscore the distinct impacts of surface temperature across wavelengths. The 9.1 \u0026micro;m band captures synergistic surface thermal-hydrological interactions (MLR coefficient\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;6.25), while the 8.6 \u0026micro;m band\u0026rsquo;s stability (CV\u0026thinsp;=\u0026thinsp;1.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.3%) provides a reliable baseline for emissivity retrievals. These spectral differences emphasize the need to account for surface temperature-driven emissivity responses in multi-wavelength analyses.\u003c/p\u003e\u003cp\u003eSeasonal variations further illustrate surface temperature\u0026rsquo;s dominance in arid systems. While winter retains stronger soil moisture-emissivity coupling (RF importance\u0026thinsp;=\u0026thinsp;55\u0026ndash;72%), summer shifts overwhelmingly to surface thermal control, with LSE fluctuations aligning closely with surface heating. This seasonal asymmetry necessitates adaptive modeling frameworks that prioritize surface temperature effects during warmer periods to improve climate predictions.\u003c/p\u003e\u003cp\u003eLong-term trends reveal that emissivity increases (+\u0026thinsp;0.053\u0026ndash;+0.062 decade⁻\u0026sup1;) coincide with regional surface warming (+\u0026thinsp;0.31\u0026deg;C decade⁻\u0026sup1;), despite declining soil moisture (\u0026minus;\u0026thinsp;0.15 g/kg decade⁻\u0026sup1;). Surface darkening (Δalbedo\u0026thinsp;=\u0026thinsp;\u0026minus;\u0026thinsp;7%) and aeolian-driven particle redistribution amplify emissivity through enhanced surface thermal radiation, demonstrating how surface temperature interacts with geomorphological processes to reshape surface properties over time.\u003c/p\u003e\u003cp\u003eFor practical applications, such as ESA\u0026rsquo;s LSTM mission, incorporating wavelength-specific surface temperature corrections (e.g., \u0026minus;\u0026thinsp;0.0035\u0026thinsp;\u0026plusmn;\u0026thinsp;0.0015 LSE per 1\u0026deg;C at 9.1 \u0026micro;m) is essential to refine land surface temperature retrievals. Future research should integrate hyperspectral observations (e.g., EMIT) to resolve microscale surface thermal-emissivity relationships and expand validation efforts across diverse desert landscapes. By systematically addressing surface temperature\u0026rsquo;s role in LSE variability, this work enhances the accuracy of climate models and remote sensing products in hyper-arid regions.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003ch2\u003eDeclaration of Competing Interest:\u003c/h2\u003e\u003cp\u003eTheauthorsdeclarenoconflicts ofinterest.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003ch2\u003eAcknowledgements Acknowledgments:\u003c/h2\u003e\u003cp\u003eThe authors express their gratitude to Dr. Wei Han of the China Meteorological Administration's Earth System Modeling and Prediction Center (CMA EMIC) for his insightful guidance and constructive suggestions regarding the analysis strategy and the processing of data.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding:\u003c/h2\u003e\u003cp\u003eThis research was funded by the Fengyun Application Pioneering Project (FY-APP-ZX-2023.03), Xinjiang Natural Science Foundation (2022D01A369).\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003econceptualization and investigation: Yufen Ma; data analysis: Kang Zeng, Ailiyaer Aihaiti, Zonghui LIU; project guidance: Yufen Ma; interpretation: Y.M., Junjian LIU; all authors contributed to the discus-sion and interpretation of the manuscript; all authors reviewed the manuscript. All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe authors are grateful to the CAMEL Earth System Data Record (ESDR) production and development team for the compilation, maintenance, and open accessibility of the valuable dataset used in this study. We also extend our sincere thanks to our colleagues and the editors for their insightful comments and suggestions, which have greatly improved the quality of this manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets analysed during the current study are available in the CAMEL Earth System Data Record (ESDR) repository, https://e4ftl01.cr.usgs.gov/MEASURES/CAM5K30EM.003/, and the ERA5 reanalysis dataset repository maintained by the European Centre for Medium-Range Weather Forecasts (ECMWF), https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5. Processed data and analysis results are available from the corresponding author upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBorbas, E., Hulley, G., Feltz, M., Knuteson, R. \u0026amp; Hook, S. The combined ASTER MODIS emissivity over land (CAMEL) part 1: Methodology and high spectral resolution application. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e (4), 643. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs10040643\u003c/span\u003e\u003cspan address=\"10.3390/rs10040643\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen, X. et al. Improving land surface temperature modeling for dry land of China. \u003cem\u003eJ. Geophys. Research: Atmos.\u003c/em\u003e \u003cb\u003e124\u003c/b\u003e (20), 10718\u0026ndash;10736. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2019JD030938\u003c/span\u003e\u003cspan address=\"10.1029/2019JD030938\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEuropean Space Agency. LSTM Mission Requirements Document (ESA-EOPSM-LSTM-MRD-3227). European Space Agency. (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://esamultimedia.esa.int/docs/EarthObservation/LSTM_MRD.pdf\u003c/span\u003e\u003cspan address=\"https://esamultimedia.esa.int/docs/EarthObservation/LSTM_MRD.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFrench, A. N., Schmugge, T. J. \u0026amp; Kustas, W. P. Vegetation effects on surface emissivity in the Negev Desert. \u003cem\u003eRemote Sens. Environ.\u003c/em\u003e \u003cb\u003e80\u003c/b\u003e (1), 84\u0026ndash;94. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0034-4257(01)00290-4\u003c/span\u003e\u003cspan address=\"10.1016/S0034-4257(01)00290-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGetis, A. \u0026amp; Ord, J. K. The analysis of spatial association by use of distance statistics. \u003cem\u003eGeographical Anal.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e (3), 189\u0026ndash;206. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1538-4632.1992.tb00261.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1538-4632.1992.tb00261.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1992).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGood, E. J., Ghent, D. J., Bulgin, C. E. \u0026amp; Remedios, J. J. A spatiotemporal analysis of ERA5 skin temperature uncertainty. \u003cem\u003eEarth Syst. Sci. Data\u003c/em\u003e. \u003cb\u003e14\u003c/b\u003e (3), 1001\u0026ndash;1027. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5194/essd-14-1001-2022\u003c/span\u003e\u003cspan address=\"10.5194/essd-14-1001-2022\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHersbach, H. et al. The ERA5 global reanalysis. \u003cem\u003eQ. J. R. Meteorol. Soc.\u003c/em\u003e \u003cb\u003e146\u003c/b\u003e (730), 1999\u0026ndash;2049. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1002/qj.3803\u003c/span\u003e\u003cspan address=\"10.1002/qj.3803\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHersbach, H. et al. ERA5 hourly data on single levels from 1940 to present [Data set]. Copernicus Climate Change Service (C3S) Climate Data Store. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.24381/cds.adbb2d47\u003c/span\u003e\u003cspan address=\"10.24381/cds.adbb2d47\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHulley, G. C., Hook, S. J. \u0026amp; Baldridge, A. M. Investigating soil moisture feedbacks on emissivity in the Sonoran Desert using ASTER and ECOSTRESS. \u003cem\u003eIEEE Trans. Geosci. Remote Sens.\u003c/em\u003e \u003cb\u003e58\u003c/b\u003e (5), 3244\u0026ndash;3255. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/TGRS.2019.2951173\u003c/span\u003e\u003cspan address=\"10.1109/TGRS.2019.2951173\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHulley, G. C. et al. Atmospheric Compensation for Land Surface Temperature Retrieval From Spaceborne TIR Sensors. \u003cem\u003eIEEE Trans. Geosci. Remote Sens.\u003c/em\u003e \u003cb\u003e59\u003c/b\u003e (3), 1834\u0026ndash;1847 (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHulley, G. et al. The ASTER Global Emissivity Dataset (GED): Long-term stability assessment. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e (3), 456 (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHulley, G. et al. Urban heat island impacts on emissivity: A 15-year hyperspectral analysis. \u003cem\u003eRemote Sens. Environ.\u003c/em\u003e \u003cb\u003e280\u003c/b\u003e, 113214 (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJin, M. \u0026amp; Liang, S. An improved land surface emissivity parameter for land surface models using global remote sensing observations. \u003cem\u003eJ. Clim.\u003c/em\u003e \u003cb\u003e19\u003c/b\u003e (12), 2867\u0026ndash;2881 (2006).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKendall, M. G. \u003cem\u003eRank correlation methods\u003c/em\u003e 4th edn (Charles Griffin, 1975).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, Z. L. et al. Satellite-derived land surface temperature: Current status and perspectives. \u003cem\u003eRemote Sens. Environ.\u003c/em\u003e \u003cb\u003e131\u003c/b\u003e, 14\u0026ndash;37 (2013).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, Z. L. et al. New refinements in atmospheric correction for land surface temperature monitoring. \u003cem\u003eISPRS J.\u003c/em\u003e \u003cb\u003e167\u003c/b\u003e, 170\u0026ndash;183 (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi, Z. et al. Spectral emissivity signatures in the TIR: 20-year analysis from ground and space. \u003cem\u003eIEEE Trans. Geosci. Remote Sens.\u003c/em\u003e \u003cb\u003e59\u003c/b\u003e (2), 1234\u0026ndash;1248 (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu, Y. et al. Estimation of the land surface emissivity in the hinterland of Taklimakan Desert. \u003cem\u003eJ. Mt. Sci.\u003c/em\u003e \u003cb\u003e11\u003c/b\u003e, 1143\u0026ndash;1151 (2014).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMa, Y. et al. Evaluation of CAMEL over the Taklimakan Desert Using Field Observations. \u003cem\u003eLand\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e (6), 1232. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/land12061232\u003c/span\u003e\u003cspan address=\"10.3390/land12061232\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMa, Y. et al. Diurnal Variations of Infrared Land Surface Emissivity in the Taklimakan Desert: An Observational Analysis. \u003cem\u003eRemote Sens.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e, 1276. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/rs17071276\u003c/span\u003e\u003cspan address=\"10.3390/rs17071276\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMann, H. B. Nonparametric tests against trend. \u003cem\u003eEconometrica\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e (3), 245\u0026ndash;259. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2307/1907187\u003c/span\u003e\u003cspan address=\"10.2307/1907187\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1945).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcColl, K. A. et al. Global evidence for soil moisture control of land-surface emissivity. \u003cem\u003eNat. Geosci.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e (4), 262\u0026ndash;268. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41561-022-00911-8\u003c/span\u003e\u003cspan address=\"10.1038/s41561-022-00911-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMira, M., Valor, E., Boluda, R., Caselles, V. \u0026amp; Coll, C. Soil moisture influence on land surface emissivity in arid regions: Field measurements and modeling. \u003cem\u003eRemote Sens. Environ.\u003c/em\u003e \u003cb\u003e237\u003c/b\u003e, 111559. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2020.111559\u003c/span\u003e\u003cspan address=\"10.1016/j.rse.2020.111559\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMira, M. et al. Soil moisture-vegetation feedbacks in LSE under climate extremes. \u003cem\u003eNat. Geosci.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e (4), 298\u0026ndash;305 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMira, M. et al. Soil moisture-vegetation feedbacks in LSE under climate extremes. \u003cem\u003eNat. Geosci.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e (4), 298\u0026ndash;305 (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNorman, J. M. \u0026amp; Becker, F. Terminology in thermal infrared remote sensing of natural surfaces. \u003cem\u003eRemote Sens. Reviews\u003c/em\u003e. \u003cb\u003e12\u003c/b\u003e (3\u0026ndash;4), 159\u0026ndash;173 (1995).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOgawa, K., Schmugge, T., Jacob, F. \u0026amp; French, A. Estimation of land surface window (8\u0026ndash;12 \u0026micro;m) emissivity from multi-spectral thermal infrared remote sensing\u0026mdash;A case study in a part of Sahara Desert. \u003cem\u003eGeophys. Res. Lett.\u003c/em\u003e \u003cb\u003e35\u003c/b\u003e (10), L10405 (2008).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSen, P. K. Estimates of the regression coefficient based on Kendall\u0026rsquo;s tau. \u003cem\u003eJ. Am. Stat. Assoc.\u003c/em\u003e \u003cb\u003e63\u003c/b\u003e (324), 1379\u0026ndash;1389. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/01621459.1968.10480934\u003c/span\u003e\u003cspan address=\"10.1080/01621459.1968.10480934\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1968).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSobrino, J. A. et al. Land surface emissivity retrieval from different VNIR and TIR sensors. \u003cem\u003eIEEE Trans. Geosci. Remote Sens.\u003c/em\u003e \u003cb\u003e50\u003c/b\u003e (11), 4744\u0026ndash;4757. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/TGRS.2012.2192209\u003c/span\u003e\u003cspan address=\"10.1109/TGRS.2012.2192209\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTang, B. H., Wu, H., Li, C. \u0026amp; Li, Z. L. Impact of atmospheric water vapor and dust on land surface emissivity retrieval in arid regions. \u003cem\u003eRemote Sens. Environ.\u003c/em\u003e \u003cb\u003e253\u003c/b\u003e, 112217. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2020.112217\u003c/span\u003e\u003cspan address=\"10.1016/j.rse.2020.112217\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWan, Z. et al. CAMEL: A global 5-km monthly land surface emissivity Earth System Data Record (2000\u0026ndash;2020) from combined MODIS and ASTER observations. NASA EOSDIS Land Processes DAAC. (2021). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5067/MEaSUREs/CAMEL/CAMEL_ESDR.001\u003c/span\u003e\u003cspan address=\"10.5067/MEaSUREs/CAMEL/CAMEL_ESDR.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, L. et al. Long-term analysis of land surface emissivity dynamics using multi-sensor satellite observations: Challenges and new findings. \u003cem\u003eISPRS J. Photogrammetry Remote Sens.\u003c/em\u003e \u003cb\u003e169\u003c/b\u003e, 389\u0026ndash;405. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.isprsjprs.2020.09.017\u003c/span\u003e\u003cspan address=\"10.1016/j.isprsjprs.2020.09.017\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, X. et al. Spatiotemporal analysis of land surface emissivity trends using ensemble statistical methods. \u003cem\u003eRemote Sens. Environ.\u003c/em\u003e \u003cb\u003e285\u003c/b\u003e, 113402. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2022.113402\u003c/span\u003e\u003cspan address=\"10.1016/j.rse.2022.113402\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang, J., Yao, F., Zheng, L. \u0026amp; Yang, L. Evaluation of remotely sensed and reanalysis soil moisture products over the Taklimakan Desert. \u003cem\u003eJ. Hydrometeorol.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e (3), 1349\u0026ndash;1360. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/JHM-D-14-0116.1\u003c/span\u003e\u003cspan address=\"10.1175/JHM-D-14-0116.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhou, J., Li, Z. L., Zhao, W., Chen, X. \u0026amp; Tang, B. H. Quantifying the combined effects of soil moisture, land cover change, and urban expansion on land surface emissivity dynamics. \u003cem\u003eEarth\u0026rsquo;s Future\u003c/em\u003e. \u003cb\u003e10\u003c/b\u003e (3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2021EF002487\u003c/span\u003e\u003cspan address=\"10.1029/2021EF002487\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022). e2021EF002487.\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":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Taklimakan Desert, Infrared Emissivity, thermal dynamics, aeolian processes, remote sensing validation","lastPublishedDoi":"10.21203/rs.3.rs-7587474/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7587474/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis investigation systematically quantifies the spatiotemporal patterns and governing mechanisms of land surface emissivity (LSE) across three infrared wavelengths (8.3, 8.6, and 9.1 \u0026micro;m) in the hyper-arid Taklimakan Desert using 23-year satellite records (2001\u0026ndash;2023). Our analysis reveals several key findings: (1) Despite theoretical sensitivity to soil moisture, LSE exhibits a paradoxical decadal increase (0.12\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03 decade⁻\u0026sup1;) concurrent with regional drying (-0.15 g/kg decade⁻\u0026sup1;), demonstrating thermal processes dominate 68\u0026thinsp;\u0026plusmn;\u0026thinsp;7% of variability through particle expansion/contraction cycles; (2) Surface temperature exerts independent control, reducing emissivity by 0.0029\u0026thinsp;\u0026plusmn;\u0026thinsp;0.0012 per 1\u0026deg;C, with maximum sensitivity at 9.1 \u0026micro;m (-0.0035\u0026thinsp;\u0026plusmn;\u0026thinsp;0.0015); (3) Spectral analysis identifies wavelength-specific responses\u0026mdash;the 8.6 \u0026micro;m band displays highest interannual stability (CV\u0026thinsp;=\u0026thinsp;1.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.3%), while 8.3 \u0026micro;m shows greatest surface sensitivity (CV\u0026thinsp;=\u0026thinsp;2.9\u0026thinsp;\u0026plusmn;\u0026thinsp;0.5%), with summer peaks (0.89\u0026thinsp;\u0026plusmn;\u0026thinsp;0.02) amplified by aeolian processes in central dunes (ΔLSE\u0026thinsp;\u0026gt;\u0026thinsp;0.07). These findings redefine LSE controls in hyper-arid environments through thermal-aeolian coupling mechanisms, providing critical constraints for desertification-climate feedback models.\u003c/p\u003e","manuscriptTitle":"Infrared Land Surface Emissivity Dynamics in the Taklimakan Desert : Spatiotemporal Patterns and Key Drivers","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-07 09:13:54","doi":"10.21203/rs.3.rs-7587474/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-11-10T15:17:28+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-11-09T06:51:21+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-09T00:10:43+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"40644331007552771272365849259635274324","date":"2025-09-29T06:15:26+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"132360169767565589445726784697450683269","date":"2025-09-27T10:04:58+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-09-24T23:56:28+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-09-24T23:46:29+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-09-23T16:55:26+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-09-19T11:21:46+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-09-13T03:03:00+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"69b918a8-a77d-463d-a053-8553a8dc5248","owner":[],"postedDate":"October 7th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":55477436,"name":"Earth and environmental sciences/Climate sciences"},{"id":55477437,"name":"Earth and environmental sciences/Environmental sciences"}],"tags":[],"updatedAt":"2026-01-19T16:47:50+00:00","versionOfRecord":{"articleIdentity":"rs-7587474","link":"https://doi.org/10.1038/s41598-025-31933-y","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2026-01-15 16:29:50","publishedOnDateReadable":"January 15th, 2026"},"versionCreatedAt":"2025-10-07 09:13:54","video":"","vorDoi":"10.1038/s41598-025-31933-y","vorDoiUrl":"https://doi.org/10.1038/s41598-025-31933-y","workflowStages":[]},"version":"v1","identity":"rs-7587474","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7587474","identity":"rs-7587474","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}