Landsat Land Surface Temperature Research Articles

Field scale partitioning of Landsat land surface temperature into soil and canopy components for evapotranspiration assessment using a two-source energy balance model

Field scale partitioning of Landsat land surface temperature into soil and canopy components for evapotranspiration assessment using a two-source energy balance model

High-resolution greenspace dynamic data cube from Sentinel-2 satellites over 1028 global major cities

Greenspace, offering multifaceted ecological and socioeconomic benefits to the nature system and human society, is integral to the 11th Sustainable Development Goal pertaining to cities and communities. Spatially and temporally explicit information on greenspace is a premise to gauge the balance between its supply and demand. However, existing efforts on urban greenspace mapping primarily focus on specific time points or baseline years without well considering seasonal fluctuations, which obscures our knowledge of greenspace’s spatiotemporal dynamics in urban settings. Here, we combined spectral unmixing approach, time-series phenology modeling, and Sentinel-2 satellite images with a 10-m resolution and nearly 5-day revisit cycle to generate a four-year (2019–2022) 10-m and 10-day resolution greenspace dynamic data cube over 1028 global major cities (with an urbanized area >100 km2). This data cube can effectively capture greenspace seasonal dynamics across greenspace types, cities, and climate zones. It also can reflect the spatiotemporal dynamics of the cooling effect of greenspace with Landsat land surface temperature data. The developed data cube provides informative data support to investigate the spatiotemporal interactions between greenspace and human society.

SSEBop Evapotranspiration Estimates Using Synthetically Derived Landsat Data from the Continuous Change Detection and Classification Algorithm

The operational Simplified Surface Energy Balance (SSEBop) model has been utilized to generate gridded evapotranspiration data from Landsat images. These estimates are primarily driven by two sources of information: reference evapotranspiration and Landsat land surface temperature (LST) values. Hence, SSEBop is limited by the availability of Landsat data. Here, in this proof-of-concept paper, we utilize the Continuous Change Detection and Classification (CCDC) algorithm to generate synthetic Landsat data, which are then used as input for SSEBop to generate evapotranspiration estimates for six target areas in the continental United States, representing forests, shrublands, and irrigated agriculture. These synthetic land cover data are then used to generate the LST data required for SSEBop evapotranspiration estimates. The synthetic LST, evaporative fractions, and evapotranspiration data from CCDC closely mirror the phenological cycles in the observed Landsat data. Across the six sites, the median correlation in seasonal LST was 0.79, and the median correlation in seasonal evapotranspiration was 0.8. The median root mean squared error (RMSE) values were 2.82 °C for LST and 0.50 mm/day for actual evapotranspiration. CCDC predictions typically underestimate the average evapotranspiration by less than 1 mm/day. The average performance of the CCDC evaporative fractions, and corresponding evapotranspiration estimates, were much better than the initial LST estimates and, therefore, promising. Future work could include bias correction to improve CCDC’s ability to accurately reproduce synthetic Landsat data during the summer, allowing for more accurate evapotranspiration estimates, and determining the ability of SSEBop to predict regional evapotranspiration at seasonal timescales based on projected land cover change from CCDC.

Filling gaps in cloudy Landsat LST product by spatial-temporal fusion of multi-scale data

Land surface temperature (LST) is an important factor in studies of surface energy fluxes between the Earth's surface and atmosphere. The Landsat LST product has been applied widely due to its fine spatial resolution and high data quality. Frequent cloud coverage, however, results in different degrees of gaps in the Landsat LST images, restricting greatly their application in practical cases requiring continuous LST time-series. Gap filling is an important technique used to reconstruct the missing data in the Landsat LST images. Using only temporally neighboring data from the same sensor, however, current gap filling methods struggle to deal with large spatial gaps and large temporal changes in LST. To solve these challenging issues, in this paper, multi-scale data (i.e., the 2 km, all-sky, hourly GOES LST product) were proposed for gap filling of Landsat LST images, and a spatial-temporal random forest (STRF) method was developed. The GOES LST product can provide effective information on LST in the gaps of the target Landsat LST data, with great potential to decreasing the uncertainty in gap filling when large LST change exists or the gap size is large. Through experiments in six regions of the USA, the proposed STRF method was shown to be more accurate than three typical gap filling and spatio-temporal fusion methods, and particularly advantageous for cases involving large temporal changes or large spatial gaps in LST. Specifically, STRF produces an average root mean square error (RMSE) of 1.6444 K and average correlation coefficient (CC) of 0.8653 for the six regions, which are 0.1133 K smaller and 0.0176 larger than the most competitive benchmark method (i.e., the spatial random forest (SRF) method). Conceptually, by spatial-temporal fusion of multi-scale data in gap filling, STRF provides a new paradigm for both spatial-temporal fusion and gap filling. Furthermore, the proposed STRF has the potential to be applied to various gap filling tasks in other situations.

Utility of thermal remote sensing for evaluation of a high‐resolution weather model in a city

AbstractProgress in high‐resolution numerical weather prediction (NWP) for urban areas will require new modelling approaches and extensive evaluation. Here, we exploit land surface temperature (LST) data from Landsat‐8 to assess 100 m resolution NWP for London (UK) on four cloud‐free days. The LST observations are directional radiometric temperatures with non‐negligible uncertainties. We consider the challenges of informative comparison between the Landsat LST and the NWP scheme's internal characterisation of the complete surface temperature. The LST spatial coverage allows large‐scale observation–model differences to be explored. In one case, obvious spatial artifacts in the NWP surface temperature are observed relative to the Landsat LST. These are found to be related to the NWP's initial method of downscaling of soil moisture using soil properties. Updated model runs have higher spatial correlation between model and Landsat LST. In cases where meteorological conditions favour the formation of horizontal convective rolls, warmer air temperatures associated with updraughts in the mixed layer extend inappropriately to the urban surface. This manifests as warm stripes in the model surface temperature that are not present in the Landsat LST. NWP–Landsat LST differences are larger in more built‐up areas on days nearer summer solstice. This is largely attributed to urban thermal anisotropy, as Landsat preferentially views warmer urban surfaces, whereas the model LST represents all surfaces. We evaluate two approaches to quantify this sampling effect, but further work is needed to fully constrain it and facilitate more informative model evaluation.

Investigation of the Efficiency of Satellite-Derived LST Data for Mapping the Meteorological Parameters in Istanbul

Land surface temperature (LST) is an essential parameter for studying environmental and ecological processes and climate change at various scales. It is also valuable for studies of evapotranspiration, soil moisture conditions, surface energy balance, and urban heat islands. Since meteorological station data can provide a limited number of point data, satellite images that provide high temporal and spatial resolution LST data in large areas are needed to be used in all these applications. In this study, the usage of satellite-derived LST images was investigated in comparison with meteorological station data measurements in Istanbul, which has heterogeneous urban structures. LST data were obtained from Landsat 5 TM, Landsat 8 OLI/TIRS, and Terra Moderate Resolution Imaging Spectroradiometer (MODIS) satellite images using the Google Earth Engine (GEE) cloud platform. The linear correlation analysis performed between Landsat LST and MODIS LST images gave a high correlation (r = 0.88). In the correlation analysis, hourly air temperature and soil temperature meteorology station data provided by the State Meteorological Service and LST values obtained from images taken from Landsat TM/TIRS and Terra MODIS were used. The correlations between air temperatures and Landsat LST ranged from 0.47–0.95 for 1987–2017 to 0.44–0.80 for MODIS LST for 2000–2017. The correlations between 5 cm soil temperatures and Landsat LST ranged from 0.76–0.93 for 2009–2017 to 0.22–0.61 for MODIS LST 2000–2017. In addition, linear regression models produced with meteorological parameters and LST values were applied to 2022 LST maps to show the spatial distribution of these parameters, and then, accuracy analyses were made.

Land Surface Temperature Retrieval From Landsat 9 TIRS-2 Data Using Radiance-Based Split-Window Algorithm

The thermal infrared sensor-2 (TIRS-2) carried on Landsat 9 is the newest thermal infrared (TIR) sensor for the Landsat project and provides two adjacent TIR bands, which greatly benefits the land surface temperature (LST) retrieval at high spatial resolution. In this article, a radiance based split window (RBSW) algorithm for retrieving LST from Landsat 9 TIRS-2 data was proposed. In addition, the split-window covariance-variance ratio (SWCVR) algorithm was improved and applied to Landsat 9 TIRS-2 data for estimating atmospheric water vapor (AWV) that is required for accurate LST retrieval. The performance of the proposed method was assessed using the simulation data and satellite observations. Results reveal that the retrieved LST using the RBSW algorithm has a bias of 0.06 K and root-mean-square error (RMSE) of 0.51 K based on validation with the simulation data. The sensitivity analysis exhibited a LST error of <1.75 K using the RBSW algorithm when the uncertainties in input parameters (i.e., AWV, emissivity, and at-sensor radiance) were considered. There is a marginal discrepancy between LST retrievals using the estimated AWV and moderate resolution imaging spectroradiometer AWV, with a difference in bias of −0.14 K and RMSE of 0.22 K, which indicates that the improved SWCVR method can provide an optional means to obtain AWV inputted in LST retrieval. With regard to the validation using the in situ measurements, the retrieved LST from Landsat 9 TIRS-2 data exhibits a bias of 0.44 K and RMSE of 1.98 K, respectively, showing a higher accuracy than the USGS Landsat LST product with bias of 1.21 and RMSE of 2.56 K. In conclusion, the proposed algorithm combining RBSW algorithm and improved SWCVR algorithm for LST retrieval from Landsat 9 has a good accuracy without dependence on external atmospheric data, and it is expected to be a reliable method for LST generation from Landsat 9 TIRS-2 data.

Application of Land Surface temperature from Landsat series to monitor and analyze forest ecosystems: A bibliometric analysis

Aim of study: Land surface temperature (LST) is an essential variable to monitor and characterize forest ecosystems. This variable has been consistently captured for almost four decades by the Landsat program. The current study aimed at identifying trends, knowledge gaps and opportunity areas in the use of Landsat derived LST for the monitoring and analysis of forest ecosystems. Materials and methods: A bibliometric analysis of scientific articles indexed in Scopus in the period 1995-2020 was conducted. Main results: Annual increase rate in the number of publications on the topic analyzed was 22.58%. The journal with more publications on the topic was Proceedings of SPIE, followed by Remote Sensing. The authors with the highest productivity on this topic were C. Quintano, I. Vorovencii, O. E. Yakubailik and M. A. Zoran. Regarding productivity by country, 38 countries with publications on this topic were identified, with the highest productivity located in China, USA and India. This group of countries also represented the most solid network of cooperation between countries. Forest ecosystems more frequently analyzed were temperate forests, followed by tropical forests. The analysis of keywords highlighted topics such as remote sensing, NDVI, MODIS and evapotranspiration. The analysis of thematic evolution indicated that areas of research and interpretation of LST data has evolved in parallel with remote sensing areas. Research highlights: Landsat LST analysis is an evolving topic with potential to contribute to improve ecosystem knowledge and to support diverse challenges in forest resources decision-making.

Analysis of Building Height Impact on Land Surface Temperature by Digital Building Height Model Obtained from AW3D30 and SRTM

Land surface temperature (LST) is heavily influenced by urban morphology. Building height is an important parameter of urban morphology that affects LST. Existing studies show contradicting results where building height can have a positive or negative relationship with LST. More studies are necessary to examine the impact of building height. However, high accuracy building height data are difficult to obtain on a global scale and are not available in many places in the world. Using the Digital Building Height Model (DBHM) calculated by subtracting the SRTM from AW3D30, this study analyzes the relationship between building height and Landsat LST in two cities: Tokyo and Jakarta. The relationship is observed during both cities’ warm seasons (April to October) and Tokyo’s cool seasons (November to February). The results show that building height and LST are negatively correlated. In the morning, areas with high-rise buildings tend to have lower LST than areas with low-rise buildings. This phenomenon is revealed to be stronger during the warm season. The LST difference between low-rise and mixed-height building areas is more significant than between mixed-height and high-rise building areas.

Reconstruction of land surface temperature under cloudy conditions from Landsat 8 data using annual temperature cycle model

Land surface temperature (LST) is an important parameter in the processes of energy exchange and water cycle between the land surface and the atmosphere. The impact of cloud cover leads to spatially incomplete of thermal infrared (TIR)-based LST products, which seriously hinders the applications of LST products in various fields. Several methods have been developed to reconstruct LST under cloudy conditions in previous studies, but there is a lack of an effective method for the reconstruction of cloudy LST at the spatial resolution of Landsat pixel (30 m). In this study, a novel method was proposed to reconstruct LST under cloudy conditions from Landsat 8 data. The LST reconstruction method includes four main steps: (1) identification of cloud-free, cloud-shadow, cloud-obscured, and cloud-covered pixels by integrating the Fmask method with a cloud-shape matching method; (2) calculation of annual temperature cycle (ATC)-based reference LST by fitting an ATC model to all available Landsat 8 LST product during 2013-2020; (3) estimation of LST residual from spatially adjacent similar pixels; and (4) estimation of reconstructed LST in terms of the sum of ATC-based reference LST and LST residual. The performance of the LST reconstruction method was evaluated using Landsat 8 LST images under clear-sky conditions as reference data. The root mean squared error (RMSE) between reconstructed LST and Landsat 8 reference LST ranges from 0.9 K to 2.5 K. The LST reconstruction method was further applied to reconstruct actual Landsat 8 LST images under cloudy conditions. Compared with original Landsat 8 LST images, the spatial distribution of reconstructed LST images is more complete. The pattern of reconstructed LST images reflects the spatial variability of LST well. The accuracy of the LST reconstruction method was validated against in situ LST measurements at six SURFRAD (Surface Radiation Budget Network) sites. The overall bias and RMSE between reconstructed LST and in situ LST at all sites are approximately −0.3 K and 3.5 K, respectively. The LST reconstruction method has great potentials to improve the applications of Landsat LST product in urban thermal environment monitoring and crop water stress monitoring.

ANALYZING IMPACTS OF LAND COVER CHANGES ON LAND SURFACE TEMPERATURE IN BAY AREA, CALIFORNIA

Abstract. Land Surface Temperature plays a critical role in the land surface-atmosphere exchange processes. In this study, the impacts of land cover changes on land Surface Temperature in bay area, California was analyzed. The key findings of the land cover change impacts study are presented. National Land Cover Database (NLCD) data and Landsat Land Surface Temperature data were used for this study. The time period of analysis is 2001 through 2020. The results of the study indicated that with the increase in built land cover, the land surface temperature in bay area, California exhibited a significant increase.

Effect of aggregation and disaggregation of land surface temperature imagery on evapotranspiration estimation

Understanding the spatial distribution of evapotranspiration (ET) and other surface energy fluxes in tree crops is vital for optimizing precision irrigation management practices. The land surface temperature (LST) is the primary input to estimate ET at field scale, but currently, available satellite data have a coarse resolution, which is not suitable for estimating ET at the required spatiotemporal scale at the field level. The development of unmanned aerial vehicles (UAVs) or airborne platforms during the last two decades has enabled the acquisition of LST and subsequent derivation of ET at the finer spatial resolution of up to a few centimeters due to the advances in low-weight thermal cameras onboard these platforms. In this study, the disaggregation methodology was applied to Landsat imagery based on aerial imagery collected at 0.5 m resolution over almond orchards in the Sacramento Valley of California, USA, to estimate ET at the orchard scale. The TsHARP algorithm was applied to sharpen the Landsat LST images (30 m spatial resolution) to target resolutions of 0.5, 10, 20 m. Also, the simple average aggregation technique was used to aggregate the aerial imagery (0.5 resolution) to 10, 20, and 30 m scales. Two Source Energy Balance (TSEB) model was used to estimate spatially distributed ET fluxes by combining the ground-based meteorological data for aggregated and disaggregated LST images. The estimated ET fluxes were compared to the measured eddy covariance fluxes, and then the aggregation and disaggregation processes were evaluated at 0.5, 10, 20, and 30 m spatial scales. At four spatial scales (0.5, 10, 20, and 30 m) aggregation and disaggregation derived ET values were highly correlated with the measured eddy covariance fluxes within the source footprint area. Furthermore, both aggregation and disaggregation approaches demonstrated similar ET spatial patterns within the study area, with slight variations observed under the Landsat 30 m resolution. The results indicated that TsHARP downscaled high-resolution imagery provides optimal spatial resolution for mapping and analyzing ET at the plant to orchard scale, with a 95% confidence interval and an RMSE of 0.43 mm/day. Overall, the results suggest that high spatial resolution imagery disaggregated from Landsat was the best alternative for estimating ET fluxes from the plant to the orchard scale, which is beneficial for precision irrigation management in orchards, but the framework can also be applied to other precision agriculture applications.

Improving LST Downscaling Quality on Regional and Field-Scale by Parameterizing the DisTrad Method

Many scientists have been investigating Land Surface Temperature (LST) because of its relevance in water management science due to its direct influence on the hydrological water cycle. This effect stems from being one of the most significant variables influencing evapotranspiration. One of the most important reasons for the evapotranspiration retrieved from MODIS data’s limited suitability for scheduling and planning irrigation schemes is the lack of spatial resolution. As a result, high-resolution LST is required for estimating evapotranspiration. The goal of this study is to improve the resolution of the available LST data, to improve evapotranspiration (ETa) estimation using statistical downscaling with Normalized Difference Vegetation Index (NDVI) as a predictor. The DisTrad (Disaggregation of Radiometric Surface Temperature) method was used for the LST downscaling procedure, which is based on aggregating the NDVI map to the LST map resolution and then calculating the coefficient of variation of the native NDVI map within the aggregated pixel and classifying the aggregated map into three classes: NDVI &lt; 0.2 for the bare soil, 0.2 ≤ NDVI ≤ 0.5 for the partial vegetation, and NDVI &gt; 0.5 for the full vegetation. DisTrad uses 25% of the pixels with the lowest coefficient of variation from each class to calculate the regression coefficients. In this work, adjustments to the DisTrad method were implemented to enhance downscaling LST and to examine the impacts of that alteration on the evapotranspiration estimation. The linear regression model was tested as an alternative to the original second-order polynomial. In using 10% of the pixels instead of the originally proposed 25% with the lowest coefficient of variation values, it is assumed that a group of pixels with a lower coefficient of variation represents a more homogeneous area, thus it gives more accurate values. The downscaled LST map retrieval was validated using Landsat 8 thermal maps (100 m). Applying the modified DisTrad approach to disaggregate Landsat LST to 30 m (NDVI resolution) yielded an R2 of 0.72 for the 10%, 0.74 for the 25% and 0.61 for the second-order polynomial lowest coefficient of variation compared to native LST Landsat, which means that 10% can be used as an alternative. Applying the downscaled LST map to estimate ETa yielded R2 0.84 in both cases, compared to ETa yielded from the native Landsat LST. These results prove that using the robust linear regression provided better results than using polynomial regression. With the downscaled Land Surface Temperature data, it was possible to create detailed ETa maps of the small agricultural fields in the test area.

Generating gapless land surface temperature with a high spatio-temporal resolution by fusing multi-source satellite-observed and model-simulated data

Land surface temperature (LST) is a key parameter when monitoring land surface processes. However, cloud contamination and the tradeoff between the spatial and temporal resolutions greatly impede the access to high-quality thermal infrared (TIR) remote sensing data. Despite the massive efforts made to solve these dilemmas, it is still difficult to generate LST estimates with concurrent spatial completeness and a high spatio-temporal resolution. Land surface models (LSMs) can be used to simulate gapless LST with a high temporal resolution, but this usually comes with a low spatial resolution. In this paper, we present an integrated temperature fusion framework for satellite-observed and LSM-simulated LST data to map gapless LST at a 60-m spatial resolution and half-hourly temporal resolution. The global linear model (GloLM) model and the diurnal land surface temperature cycle (DTC) model are respectively performed as preprocessing steps for sensor and temporal normalization between the different LST data. The Landsat LST, Moderate Resolution Imaging Spectroradiometer (MODIS) LST, and Community Land Model Version 5.0 (CLM 5.0)-simulated LST are then fused using a filter-based spatio-temporal integrated fusion model. Evaluations were implemented in an urban-dominated region (the city of Wuhan in China) and a natural-dominated region (the Heihe River Basin in China), in terms of accuracy, spatial variability, and diurnal temporal dynamics. Results indicate that the fused LST under all-weather conditions is highly consistent with actual Landsat LST data (in situ LST measurements), in terms of a Pearson correlation coefficient of 0.94 (0.96–0.99), a mean absolute error of 0.71–0.98 K (0.82–3.34 K), and a root-mean-square error of 0.97–1.26 K (1.09–4.36 K). The generated diurnal Landsat-like LSTs under all weather conditions are able to support diurnal dynamic studies that are the most relevant to human activities, such as the study of urban heat islands (UHIs) and water resource management at the field scale.

Inferring Land Conditions in the Tumen River Basin by Trend Analysis Based on Satellite Imagery and Geoinformation

The aim of this study was to map the land condition within the area of the Tumen River Basin (TRB), located on the Sino–North Korean border, using trend analysis of environmental factors. The normalized difference vegetation index (NDVI) and land surface temperature (LST) trends over the past 30 years were analyzed to identify areas that have undergone degradation, restoration, and/or a transition. Landsat NDVI and LST were obtained using the Google Earth Engine (GEE) platform. Erosion was also gauged over the same period using the Revised Universal Soil Loss Equation (RUSLE). Our results showed that only 0.3% of the land within the TRB underwent change that can be characterized as statistically significant within the study period. We therefore infer that land degradation may not be a major concern in the study area. Areas with a significant upward trend of soil loss accounted for 0.8% of the basin’s footprint and were mainly distributed upstream of North Korea. However, more than 80% of the area was found to be suffering from water stress, 10% of these areas were statistically significant and most were located downstream.

Long‐Term and Fine‐Scale Surface Urban Heat Island Dynamics Revealed by Landsat Data Since the 1980s: A Comparison of Four Megacities in China

AbstractLong‐term and fine‐scale monitoring of surface urban heat island (SUHI) is critical for the design of heat mitigation strategies. Landsat series offer long‐term (since the 1980s) and fine‐scale land surface temperature (LST) observations for such SUHI analysis. However, Landsat data are characterized by a long revisit period (16 days) and are seriously impacted by cloud contamination and stripe gaps, making the Landsat‐derived long‐term SUHI trends across cities incomparable. To address this issue, here we applied the Prophet model to reconstruct temporally consistent long‐term (∼1985–2019) clear‐sky Landsat LSTs over four megacities in China (Beijing, Chongqing, Shanghai, and Shenzhen). The results show that the mean absolute error of the SUHI intensity (SUHII) estimated using the reconstructed LSTs ranges from 0.2 to 0.6°C. The reliability of the reconstructed LSTs is further evidenced by the general consistency between the reconstructed Landsat and original Moderate Resolution Imaging Spectroradiometer LSTs. Our analysis further demonstrates that the overall SUHII has been increasing slowly over Beijing but has been increasing rapidly in the other three megacities since the 1980s. Local SUHII trends also varied between urban regions. Urban core is characterized by an initial increase and then a slight decrease in the local SUHII, while a continuously increasing trend is observed for the urban fringe. For the new urban development, the interannual SUHII trend lies somewhere between these trends. Overall, our study provides a practical approach to reconstruct long‐term clear‐sky Landsat LSTs, and delivers a better understanding of long‐term and fine‐scale SUHI variations across cities.

Burned Area Detection Using Multi-Sensor SAR, Optical, and Thermal Data in Mediterranean Pine Forest

Burned area (BA) mapping of a forest after a fire is required for its management and the determination of the impacts on ecosystems. Different remote sensing sensors and their combinations have been used due to their individual limitations for accurate BA mapping. This study analyzes the contribution of different features derived from optical, thermal, and Synthetic Aperture Radar (SAR) images to extract BA information from the Turkish red pine (Pinus brutia Ten.) forest in a Mediterranean ecosystem. In addition to reflectance values of the optical images, Normalized Burn Ratio (NBR) and Land Surface Temperature (LST) data are produced from both Sentinel-2 and Landsat-8 data. The backscatter of C-band Sentinel-1 and L-band ALOS-2 SAR images and the coherence feature derived from the Interferometric SAR technique were also used. The pixel-based random forest image classification method is applied to classify the BA detection in 24 scenarios created using these features. The results show that the L-band data provided a better contribution than C-band data and the combination of features created from Landsat LST, NBR, and coherence of L-band ALOS-2 achieved the highest accuracy, with an overall accuracy of 96% and a Kappa coefficient of 92.62%.

A Scale-Separating Framework for Fusing Satellite Land Surface Temperature Products

The trade-off between spatial and temporal resolutions of satellite imagery is a long-standing problem in satellite remote sensing applications. The lack of daily land surface temperature (LST) data with fine spatial resolution has hampered the understanding of surface climatic phenomena, such as the urban heat island (UHI). Here, we developed a fusion framework, characterized by a scale-separating process, to generate LST data with high spatiotemporal resolution. The scale-separating framework breaks the fusion task into three steps to address errors at multiple spatial scales, with a specific focus on intra-scene variations of LST. The framework was experimented with MODIS and Landsat LST data. It first removed inter-sensor biases, which depend on season and on land use type (urban versus rural), and then produced a Landsat-like sharpened LST map for days when MOIDS observations are available. The sharpened images achieved a high accuracy, with a RMSE of 0.91 K for a challenging heterogeneous landscape (urban area). A comparison between the sharpened LST and the air temperature measured with bicycle-mounted mobile sensors revealed the roles of impervious surface fraction and wind speed in controlling the surface-to-air temperature gradient in an urban landscape.

Improving land surface temperature prediction using spatiotemporal factors through a genetic-based selection procedure (Case Study: Tehran, Iran)

In recent decades, land surface temperature (LST) has been considered as an important satellite product for thermal monitoring of the environment. Despite the great development of LST retrieval technology from satellite data, especially from Landsat images, up-to-date thermal information cannot always be obtained. Predicting LST values for future times is a pressing issues to understand and mitigate the effects of climate change, especially in urban environments. This highlights the importance of LST forecasting methods. Accordingly, in this study a combination of Genetic Algorithm and Geographically and Temporally Weighted Regression (GTWR) is presented to predict LST based on nine selected spatiotemporal factors affecting LST in the urban areas. The genetic algorithm is used to determine weights of different factors in the LST prediction with respect to the GTWR as the regression model. Then, these weights are used to obtain predicted LST values for experimental points. Landsat-derived LST bands is used to train the genetic algorithm and extract the factors’ weights. Tehran city in the summer is selected as the case study in our work. Landsat LST data from 2013 to 2016 are used for training. To assess the performance of our scheme, we apply it in 2017. We find a correlation coefficient of 0.99 and a root-mean-square error of 0.5 K with respect to the Landsat data. The results of this study also indicate that LST prediction using weighted factors is superior to using all factors with equal contributions.

Spatiotemporal Temperature Fusion Based on a Deep Convolutional Network

High-spatiotemporal-resolution land surface temperature (LST) images are essential in various fields of study. However, due to technical constraints, sensing systems have difficulty in providing LSTs with both high spatial and high temporal resolution. In this study, we propose a multi-scale spatiotemporal temperature-image fusion network (MSTTIFN) to generate high-spatial-resolution LST products. The MSTTIFN builds nonlinear mappings between the input Moderate Resolution Imaging Spectroradiometer (MODIS) LSTs and the out- put Landsat LSTs at the target date with two pairs of references and therefore enhances the resolution of time-series LSTs. We conduct experiments on the actual Landsat and MODIS data in two study areas (Beijing and Shandong) and compare our proposed MSTTIFN with four competing methods: the Spatial and Temporal Adaptive Reflectance Fusion Model, the Flexible Spatiotemporal Data Fusion Model, a two-stream convolutional neural network (StfNet), and a deep learning-based spatiotemporal temperature-fusion network. Results reveal that the MSTTIFN achieves the best and most stable performance.