High-resolution leaf area index (LAI) retrieval is crucial for ecological and agricultural applications, yet it remains challenging due to the enhanced spatial heterogeneity and limited temporal observations of high-resolution imagery. Although combining physics-based models with deep learning algorithms is a promising approach, existing approaches are limited by the restricted application scope of physical models and the lack of deep learning networks that can capture both spatial and temporal information. This study develops an image-to-image algorithm, termed UM-RCA (unified bidirectional reflectance distribution function model-informed ResNet-ConvLSTM-attention), that integrates a physical model with a new spatiotemporal deep learning network to generate 30-m LAI products using Landsat imagery. The method first applies a unified bidirectional reflectance distribution function model, suitable for both continuous and discrete canopies, to generate a high-fidelity training dataset. A ResNet-ConvLSTM-Attention (RCA) network is then trained on the dataset to learn the relationships between Landsat reflectance and LAI patch time series. The three components of the RCA network combine spatial, temporal, and spectral information through an "Extract-Aggregate-Refine" structure. Results show that the UM-RCA algorithm achieves high accuracy on LAI field measurements across different biome types (overall MAE = 0.510, RMSE = 0.685) and produces more spatiotemporally coherent LAI maps than three mainstream algorithms or products: the random forest model, the Sentinel application platform biophysical processor, and the High-resolution Global LAnd Surface Satellite LAI product. The UM-RCA algorithm offers significant practical value for precision agriculture and ecological monitoring.

Land surface temperature (LST) serves as a key parameter for analyzing surface energy balance and hydrological cycles, driving the need for high-accuracy, high-spatial-resolution LST products in Earth system science. The wide thermal infrared imager (WTI) onboard China's GF-5 01A satellite provides four-channel thermal infrared (TIR) data with a wide swath (over 1500 km), high-spatial resolution (better than 100 m at nadir), and a two-day revisit cycle, integrating the extensive coverage (like MODIS and VIIRS) with the high-spatial-resolution characteristics (like Landsat-8/9). This study developed multiple LST retrieval approaches for WTI data, including the temperature-emissivity separation (TES) algorithm, split-window (SW) algorithm, and the hybrid algorithm. Three kinds of algorithms demonstrated promising accuracy in simulation experiments, and were applied to real WTI images collected throughout the year of 2024 in different seasons over China. Validation against ground measurements from the HiWATER network showed that a three-channel hybrid model algorithm, TSW3T, achieved optimal performance, with an overall RMSE of 2.46 K and bias of -0.15 K. Four specific sites (DSL, HHZ, SDZ, and YKZ) maintained RMSE values below 2.0 K, demonstrating satisfactory accuracy. Specifically, at high-altitude sites, the three-channel algorithms consistently outperformed their four-channel counterparts. The TSW3T algorithm delivered the best validation results, achieving an RMSE of 2.22 K. Of the selected 93 matching points, the cross-comparison between TSW3T-derived LST and MODIS LST products showed an RMSE of 1.59 K and a bias of -0.17 K under an average overpass time difference of 22 min between WTI and MODIS. These findings confirm the robustness of the TSW3T for generating high-spatial-resolution LST and emissivity products from GF-5 01A WTI TIR data.

The thermal radiation directionality (TRD) effect is a critical factor influencing the global application of thermal infrared remote sensing data, and limiting the comparison of long-term temporal data series. The inherent inconsistency of thermal radiation data acquired from different observation angles poses significant challenges for data utilization and analysis. To address this issue, this study develops a dual-band parametric angular normalization (DPAN) method based on physical model. The key parameters for angular normalization are firstly derived through theoretical analysis, then empirical estimation formulas for these parameters are established. Analysis based on simulation dataset demonstrates that the DPAN method can correct an average of 62 % of thermal radiation angular effects, with maximum corrections reaching 66-68 % and minimum corrections approximately 47-49 % under various input parameter errors. A wavelength difference of 2-3.5 mu m is found to be the best for the dual-band setting to apply the DPAN method. When applied to actual Moderate Resolution Imaging Spectroradiometer (MODIS) data, the method achieves a radiative brightness temperature root mean square difference (RMSD) of approximately 1.3 K, with maximum correction exceeding 8 K. Notably, temperature corrections typically surpass 1 K when the view zenith angle exceeds 40 degrees. In summary, the DPAN method requires neither multi-angle nor multi-temporal observations, enabling direct application to any satellite dataset with at least two thermal infrared bands. These characteristics satisfy the essential requirements for operational angular normalization of satellite-derived land surface temperature products, indicating strong potential for widespread implementation.

China is one of the most severely desertified countries in the world. Currently, the use of topsoil characteristics in the thermal infrared bands is limited due to the lack of thermal infrared remote sensing products. However, the Combined Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) and MODIS Emissivity over Land (CAMEL) product has the potential for sandy land detection. Therefore, in this study, optical reflectance data from the Sentinel-2 dataset and thermal emissivity data from the CAMEL dataset were combined to estimate the sand content of the Mu Us sandy land, which is located in the semi-arid and arid regions of northwest China. Four sand indicators were selected to represent the spectral information. The sample dataset was established on basis of the ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS) spectral library. Three machine learning regression methods were used to identify the relationship between the spectral information and the sand content of the samples. Random forest regression demonstrated the best performance, with a RMSE of 0.027 for the training set and a RMSE of 0.074 for the test set. The sand content map of the Mu Us sandy land, generated using the pre-trained random forest regression model, suggests that the sand content was approximately 35%. The new method is a promising way to track changes in sand content over large areas from time-series remote sensing data.

Multi-angle thermal infrared (MATIR) remote sensing provides valuable day-night and multiple angular information that is of significant practical value in applications. However, multi-angle data heterogeneity is one of the core challenges in feature learning and scene understanding, which could severely degrade the model inference performance of deep neural networks. To address this issue, this study proposes a new fine-grained dataset and a unified method for the MATIR object detection task. In detail, the fine-grained MATIR object detection (MATIROD) dataset is captured by an unmanned aerial vehicle (UAV)-based platform, which offers significant advantages in terms of cost efficiency and exceptional maneuverability. The MATIR-OD dataset comprises 24 finegrained and multi-angle data subsets, containing a total of 43,540 instances. Moreover, the unified MATIR object detection method, denoted as U-MATIR, includes the heterogeneous label space module and hybrid view cascade module. In the multi-angle object detection task, based on four public datasets and the proposed dataset, the all-angle experimental results show that the U-MATIR outperforms the ground- or aerial-view object detection models, increasing accuracy with an approximately 18-65% improvement in the mean Averaged Precision (mAP) metric, which exhibits notable robustness and generalization ability. In addition, the extensive experiments demonstrate the boundaries of robustness and generalization ability under 20-120 m and 30-90 degrees fine-grained observation data. In particular, the optimal detection angle is defined as 60 degrees under the above observation heights. The MATIR object detection dataset and unified method provide new insight for accurate multi-angle localization and achieve competitive detection performance.

Land Surface Temperature(LST)is a critical parameter for understanding surface energy balance and water cycle processes.This study aimed to develop LST retrieval algorithms using thermal infrared data from the Visual and Infrared Multispectral Imager(VIMI)onboard Chinese Gaofen-5B(GF-5B)satellite,providing a new data source for global high-spatial-resolution LST product generation. A simulated dataset was first constructed on basis of the conventional thermal infrared radiative transfer model,TIGR-3 atmospheric profiles,and 74 typical surface emissivity samples selected from the ASTER spectral library.Two split-window algorithms(called as SW-1 and SW-2)for GF-5B thermal infrared data were developed based on the simulated dataset and sensor characteristics,with both algorithms achieving an RMSE of 1.13 K.Considering the influence of Column Water Vapor(CWV)on LST retrieval,algorithm coefficients were derived under different CWV conditions.Comparison of RMSE under identical conditions demonstrated that SW-2 generally outperformed SW-1 in accuracy.The algorithm's performance varies with CWV,with RMSE values of SW-1(SW-2)decreasing from 1.41 K(1.34 K)in high CWV conditions(4.5-6.3 g/cm2)to 0.40 K(0.39 K)in low CWV conditions(0.0-2.5 g/cm2).The SW-2 algorithm was thus recommended as the preferred choice for GF5B-based LST retrieval applications.Key parameters acquisition methods were also proposed,including land surface emissivity(via a modified NDVI-NDWI threshold method)and atmospheric CWV(via a water vapor split-window covariance-variance ratio method). The proposed algorithm was applied to four experimental regions with different land cover types.Combined with false-color composite imagery(green,red,and near-infrared bands)derived from VIMI's visible and near-infrared data,the spatial distribution of LST retrieval results was evaluated in relation to surface characteristics.Artificial surfaces and bare soil exhibited relatively higher LST values,while water bodies and vegetation showed lower LST values,consistent with thermal radiative characteristics.Validation was conducted using in-situ LST measurements from HiWATER sites and MODIS LST products.Results demonstrated that the LST errors for daytime and nighttime retrievals were 1.88 K and 0.99 K,respectively.Among all validated data,the maximum daytime temperature difference reached 2.86 K,while the nighttime maximum difference was 1.27 K.Additionally,variations in validation accuracy were observed across imagery acquired at different periods for the same site.For cross-comparison with MODIS MOD1 1 Al LST products,observational discrepancies were minimized by aligning acquisition times and viewing angles,and spatial aggregation was applied to reduce the impact of resolution mismatches.High-resolution Sentinel-2 land cover products and temperature gradients between adjacent pixels were utilized as screening criteria.Except for the Tangshan(Hebei)region,which exhibited significant deviations due to cloud contamination,temperature differences in the remaining three regions remained within 1.5 K. The LST retrieval results across the four experimental regions exhibited reasonable spatial distributions and strong correlations with land cover types.Validation with limited in-situ HiWATER data confirmed higher accuracy for nighttime retrievals compared to daytime.Cross-comparison with MOD1 1 Al LST products demonstrated acceptable agreement(differences ≤1.5 K)in three regions,indicating that the proposed algorithm achieves satisfactory accuracy.

The observation of geostationary orbit satellites,with the characteristics of wide range,high frequency,and fixed point observation,provides an important way to obtain land surface and atmospheric parameters and can monitor the change in land surface temperature over long time series.As the second generation of China's geostationary meteorological satellites,FY-4 A and B satellites'constellation observation greatly expand the scope of meteorological observation and improve data utilization efficiency.Before the joint use of the two satellites'data,the radiometric consistency between the same-band observations of the two satellites needs to be explored. Taking the thermal infrared band data as an example,this study assessed the radiometric consistency between three thermal infrared bands(centered at 8.5 μm,10.8 μm,and 12.0 μm)of FY-4 A and B satellites in four experimental areas:Dunhuang calibration field,Hulunbuir Grassland,Chaohu Lake,and South China Sea.The heterogeneity of the Dunhuang calibration field and Hulunbuir Grassland study areas was evaluated first.A method was then proposed to correct the angular effects and spectral response function differences in observation from geostationary orbit satellites.This method corrected the angular effects by establishing empirical relationships between simulated radiance data at different viewing angles and eliminated spectral response function differences by correcting brightness temperature and radiance on the basis of lookup tables.Finally,this method was used to correct the thermal infrared data of the two satellites at the same observation time,and the brightness temperatures after correction were compared to analyze the radiometric consistency of the corresponding thermal infrared bands. Based on the data analysis,after the removal of the differences in space-time,observation angle,and spectral response function of the two satellites,the results show a strong positive correlation between the brightness temperatures of the three thermal infrared bands on FY-4 A and B satellites.The brightness temperature errors are small,indicating good radiometric consistency.However,a slight variation in brightness temperature exists among those different thermal infrared bands.The consistency of brightness temperature of the second thermal infrared band is better than that of the third band,and the third band performs better than the first band.The root-mean-square errors of brightness temperatures for the three bands range from 0.28 K to 1.51 K,with deviations between-1.13 K and 0.85 K.The brightness temperature deviations in the second and third bands exhibit a noticeable positive skewness,whereas the deviations in the first band show a negative skewness. Comparison of the brightness temperature data before and after correction suggests that the method proposed in this paper has good applicability for the study of radiometric consistency in thermal infrared bands of geostationary orbit satellites.The results show that the radiometric consistency of thermal infrared radiation between the two satellites is generally good,although it might be influenced by land cover types.The findings presented in this paper provide important guidance for the joint utilization of thermal infrared data from FY-4 A and B satellites.

Urban resilience is crucial for achieving a city’s sustainable development goals in the context of global change. However, a lack of sensing and intelligent analytical frameworks that can monitor the fine-grained spatiotemporal dynamics of the stressors and a city’s reactions to these stressors is largely prohibiting the smart decision support for urban resilience studies. Here, we propose the integration of multi-source sensing technologies, including remote sensing, social sensing, and IoT-based sensing, collectively referred to as “fusion sensing,” to enhance our monitoring capability of urban systems. These technologies enable real-time data collection and analysis, facilitating adaptive responses to disturbances. Additionally, the artificial intelligence methods play important roles in both short-term emergency management and long-term planning, optimizing resource allocation and resilience. Hence, we call for collaboration among governments, industry, academia, and individuals to leverage artificial intelligence and various sensing technologies, thereby enhancing urban resilience through proactive and informed decision-making.

Thermal infrared(TIR)remote sensing is a branch of remote sensing that focuses on the acquisition,processing,and interpretation of data in the TIR region of the electromagnetic spectrum,ranging from 3 μm to 15 μm.This technology is critical for studying the thermal radiation characteristics of the Earth's surface and its changes.Over the last 6 decades,remarkable achievements have been achieved in sensor development,theoretical understanding,and methodological approaches within TIR remote sensing.At present,more than 60 operational satellites are equipped with TIR sensors,thereby offering substantial improvements in spatial resolution,detection channels,and capabilities.For example,the thermal bands of the widely used Moderate Resolution Imaging Spectroradiometer(MODIS)provide a spatial resolution of 1 km;however,the recent HotSat-1,launched in June 2023,offers the highest-resolution commercial TIR sensor in orbit,capable of identifying features as small as 3.5 m.This progress marks a new era in Earth observation and climate monitoring.Moreover,hyperspectral TIR data offers unique insights into the Earth's surface and atmosphere. Alongside sensor advancements,notable progress has been made in the retrieval of atmospheric and surface parameters.Various methods,such as the single-channel,split-window,temperature,and emissivity separation,as well as the day-night methods,have been developed to derive Land Surface Temperature(LST)accurately.LST is a critical parameter in the land-atmosphere energy balance.Advancements in TIR observation retrieval methods have driven the development of various LST products.At present,more than 30 types of LST products are publicly available,including those from Landsat,the Advanced Spaceborne Thermal Emission and Reflection Radiometer,MODIS,the Advanced Baseline Imager,and the Satellite Application Facility on Land Surface Analysis.The integration of physics-based models with data-driven Artificial Intelligence(AI),the"TIR+AI"paradigm,can considerably enhance the interpretation of TIR images and information extraction capabilities. These innovations have expanded TIR's applications across diverse fields,including natural resource monitoring,ecological assessment,disaster response,planetary exploration,human health evaluation,and public safety.For example,security agencies typically rely on fingerprints and documentation to verify individuals' identities.Moreover,current research indicates that analyzing skin temperatures with infrared cameras makes detecting fraudulent identities possible with an acceptable margin of error.Furthermore,TIR remote sensing is expected to play increasingly important roles in the exploration of deep space,thereby providing insights into planetary components,mineral exploration,space weather,and human habitation.Therefore,the benefits of TIR are significant and have contributed significantly to national economies and the societal well-being of individuals and populations. This study presents a comprehensive review of the state of the art in the field of TIR remote sensing by discussing the historical development of TIR sensors,the evolution of retrieval methods,and various applications.It also explores future trends and challenges,thereby highlighting opportunities for further innovation.In the coming years,several key technologies,namely,advanced sensor systems,AI integration with physical models,real-time remote sensing,and data fusion with optical and microwave sources,are expected to enhance the understanding and management of environmental,geological,anthropogenic,and disaster phenomena,thereby positioning TIR remote sensing as an indispensable tool for researchers,policymakers,and emergency responders.

This study monitors four major industrial parks in Beijing using thermal infrared remote sensing data from Landsat 8 and 9. Land surface temperature (LST) was retrieved through a split-window algorithm, and two key metrics, temperature deviation and temperature anomaly, were defined to analyze thermal activities in key areas. The monthly LST variations in 2024 followed a consistent seasonal pattern, with higher temperatures in summer and lower in winter. High-temperature areas in JYBS, JYLS, KFQ, and YH exhibited significant differences in thermal behavior, reflecting production activity levels. The dataset, updated every eight days, enabled timely monitoring and analysis. While temperature deviation proved more reliable than temperature anomaly due to reduced seasonal influences, the results highlight the need for further refinement. This research provides valuable insights for regulating industrial heat emissions and assessing their impact on urban thermal environments.

Leaf Area Index (LAI) and Leaf Chlorophyll content (LCC) are vegetation structural parameters and physiological parameters respectively, they are two critical indicators for forest growth condition and global carbon cycle. Accurately estimating LAI and LCC based on remote sensing at high spatial resolution is of great significance for forest dynamics monitoring. The APPLE-GO model is a novel forest canopy BRDF model established at high spatial resolution scenes, providing a new approach for the inversion of biophysical variables at high spatial resolution. In this study, a two-step method was used to retrieve LAI and LCC separately from forest at high spatial resolution imagery. The first step is estimating LAI using adaptive gradient descent algorithm, then based on the retrieval LAI, the leaf reflectance was converted by canopy reflectance using Look-up tables (LUTs) generated from the APPLE-GO model. The second step is retrieving LCC from derived leaf reflectance using the PROSPECT-PRO model. Retrieved LAI and LCC were validated against simulation in LESS. The results indicate a good agreement between retrieved and simulated LAI with RMSE of 0.310 and R-2 of 0.87. Retrieved LCC is of good quality compared to simulated LCC, with R-2 of 0.64 and RMSE of 4.25 mu g/cm(2). This method is also applied to retrieve the LAI and LCC of forests using BJ-3 high-resolution remote sensing images, yielding promising results. In conclusion, this study shows the potential in estimating LAI and LCC of forest using high spatial resolution imagery base on APPLE-GO model.

Nighttime light (NTL) remote sensing data have been widely used in various fields, such as human activity analysis, urbanization studies, and economic evaluation. However, Earth's nighttime environment is very complex so that the NTL images are seriously affected by atmospheric effects and moonlight. This complexity primarily stems from the numerous atmospheric scattering and absorption, as well as the incoming moonlight, which can significantly distort and contaminate NTL observed by the satellite and consequently reduce the precision and stability of NTL data. In order to improve the quantitative quality of the NTL data, this article proposes an innovative atmospheric correction algorithm that leverages the nighttime radiative transfer model (nRTM) considering both atmospheric effect and moonlight effect to get ground radiance of artificial lights from satellite NTL images. This model takes into account the complex interactions between light and the atmosphere. By simulating these processes, the algorithm is able to separate the contributions of atmospheric scattering and absorption from the original NTL images. To demonstrate the effectiveness of the proposed algorithm, this article takes the Sustainable Development Goals Science Satellite 1 (SDGSAT-1) NTL image of Beijing as a representative case study of atmospheric correction. By comparing the corrected and uncorrected images, it is evident that the atmospheric correction significantly improves the quality of the NTL data and the ground nighttime lighting information becomes clearer and more accurate, effectively removing noise interference and enhancing data reliability. Moreover, it also found that the high-pressure sodium (HPS) lamps and LED lamps in the NTL images presented different radiance values and spectral shapes that can be helpful for classifying different lamps.

Forests play a critical role in global carbon cycle and climate change. Satellite-based remote sensing observations provide an efficient way to monitor and manage forests across vast areas. In recent decades, the number of high-resolution satellites in orbit has been increasing rapidly. However, classic analytical models for forest canopies are developed at large scale, so applying them directly at the high resolution may introduce significant uncertainties, which stems from the shading effect of adjacent pixels and the influence of cross-radiation. This paper proposes a high-spatial resolution forest canopy reflectance model named APPLE-GO. The new model introduces the two-dimensional path length distribution (2-PLD) concept and shading factors to describe the shading effect of adjacent pixels, and the cross-radiation is estimated by an eight-neighborhood convolution algorithm. The bi-directional reflectance factor (BRF) calculated by the APPLE-GO model was validated by the 3D radiative transfer model LESS, with a RMSE of 0.01 in the red band and a RMSE lower than 0.06 in the NIR band, which indicates that the new model can accurately model the forest canopy reflectance at high-spatial resolution.

Urban heat islands and frequent heatwave exacerbate the risk of heat-related morbidity and mortality. This study demonstrates the significant and previously underestimated influence of the adjacency effect in high-resolution studies of the urban thermal environment. The contribution of three-dimensional (3D) geometry of target and adjacent landscapes on thermal environment were simulated by leveraging the urban thermal radiative transfer (UCM-RT) model. Typical two-dimensional and 3D metrics were calculated to examine the relationships between characteristics of urban landscapes and land surface temperature (LST). Our results demonstrate a threefold stronger thermal impact from adjacent 3D urban landscapes compared to internal landscapes (maximum thermal effects: 3.8 K vs. 0.7 K), emphasizing their crucial role in shaping thermal patterns. The bivariate correlation analysis has also revealed that the correlation between neighboring 3D landscape metrics and urban LST is higher by 0.09 than that of the inside landscape. A strong positive relationship exists between the 3D landscape metrics with urban LST. The results highlight the significance of adjacent thermal effects in refined urban thermal environment studies. These findings support the in-depth exploration of urban thermal processes to minimize the potential health impacts of thermal risk.

Forests are the key component of terrestrial ecosystems, playing a vital role in the global carbon and water cycles as well as in climate change. Satellite remote sensing imagery has the advantage of quantitatively monitoring and assessing the health status of forest canopies at large scales. With the improvement in spatial resolution of satellite sensors, it has become feasible to conduct quantitative research at high spatial resolutions (< 10 m). However, classic physical models that are based on simplified assumptions and only account for the radiative transfer process within the target pixel face challenges in supporting quantitative analysis at high-resolution scales, as high-resolution pixels are subject to significant radiative influences from adjacent pixels. In this study, we propose a high-spatial resolution forest canopy reflectance model, APPLE-GO, which comprehensively considers the shading effect and cross-radiation caused by adjacent pixels. The two-dimensional path length distribution (2-PLD) method is used to calculate the area fractions of each component, while shading factors are introduced to quantitatively calculate the reductions in the area fractions of sunlit components due to adjacent pixels. Multiple scattering energy is calculated based on the spectral invariant theory and the eight-neighborhood convolution algorithm. The bi-directional reflectance factor (BRF) calculated by the APPLE-GO model was evaluated against the three-dimensional (3D) radiative transfer model LESS, yielding RMSEs/RRMSEs of 0.008/10.2 % and 0.054/15.9 % in the red and near-infrared (NIR) bands, respectively. The model was also validated with satellite observations, showing RMSEs below 0.01 (RRMSE <27 %) for larch forests and under 0.017 (RRMSE <35 %) for mixed forests in the visible bands. These results demonstrate that the proposed model can accurately calculate the BRF in the nadir viewing direction, highlighting its potential for extracting vegetation parameters from high-resolution remotely sensed imagery.

As a pivotal physiological trait influencing a plant's photosynthetic capacity, accurate and efficient characterizing of leaf chlorophyll content (LCC) is crucial for understanding terrestrial ecosystem carbon cycling. The rededge reflectance spectrum, obtained through remote sensing, offers valuable insights for LCC estimation. The rededge position (REP) in the vegetation spectral reflectance is a potent proxy for LCC due to its relationship with chlorophyll absorption. However, variations in leaf area index (LAI) can influence the REP and impact its performance in LCC estimation. In this study, we propose the Sentinel-2 leaf chlorophyll index (S2LCI), a novel vegetation index based on REP and a LAI indicator using Sentinel-2 multispectral red-edge bands. This innovative index significantly mitigates the influence of LAI variation and enhances the LCC. The effectiveness of S2LCI has been evaluated through multiple ways, including PROSAIL simulated datasets, ground-measured LCC with canopy spectra, and Sentinel-2 imagery. Our results demonstrate strong agreement between S2LCI and LCC in both simulated and ground-measured datasets (R = 0.492 for ground spectra and R = 0.526 for Sentinel-2 imagery), outperforming classic LCC-related vegetation indices. Furthermore, S2LCI mapping reveals finer spatial details compared to LCC derived using the Sentinel Application Platform (SNAP) biophysical processor. This study highlights the suitability of S2LCI for LCC estimation and offers a promising solution, complementing other LCC retrieval approaches to rapidly generate decameter-scale crop LCC maps using Sentinel-2 imagery.

The estimation of Sea Surface Temperature (SST) from ocean satellites with large observation angles must account for the angular effects on SST. This study developed an angledependent non-linear split-window algorithm (A-NLSW) to retrieve SST from Chinese ocean satellite HY-1D thermal infrared data. The algorithm coefficients were obtained based on the simulated dataset and grouped by initial SSTs, total atmospheric column water vapor content (TCWV), and satellite zenith angle (SZA). The A-NLSW algorithm is validated and re-calibrated using the bulk temperature collected by the iQuam in-situ dataset. After re-calibration, the accuracy of the SST was improved from 1.53 K to 0.87 K for SZA ranging from 0 to 70.5 degrees. Nearly 60% of the validation points achieved an accuracy of 0.5 K and over 90% achieved an accuracy of 1.0 K. These findings highlight the robustness of the A-NLSW algorithm in reliably retrieving SST from HY-1D satellite images, even when observations are made at large SZA.