Investigation and validation of algorithms for estimating land surface temperature from Sentinel-3 SLSTR data

Land surface temperature (LST) is an important indicator of global ecological environment and climate change. The Sea and Land Surface Temperature Radiometer (SLSTR) onboard the recently launched Sentinel-3 satellites provides high-quality observations for estimating global LST. The algorithm of the official SLSTR LST product is a split-window algorithm (SWA) that implicitly assumes and utilizes knowledge of land surface emissivity (LSE). The main objective of this study is to investigate alternative SLSTR LST retrieval algorithms with an explicit use of LSE. Seventeen widely accepted SWAs, which explicitly utilize LSE, were selected as candidate algorithms. First, the SWAs were trained using a comprehensive global simulation dataset. Then, using simulation data as well as in-situ LST, the SWAs were evaluated according to their sensitivity and accuracy: eleven algorithms showed good training accuracy and nine of them exhibited low sensitivity to uncertainties in LSE and column water vapor content. Evaluation based on two global simulation datasets and a regional simulation dataset showed that these nine SWAs had similar accuracy with negligible systematic errors and RMSEs lower than 1.0 K. Validation based on in-situ LST obtained for six sites further confirmed the similar accuracies of the SWAs, with the lowest RMSE ranges of 1.57–1.62 K and 0.49−0.61 K for Gobabeb and Lake Constance, respectively. While the best two SWAs usually yielded good accuracy, the official SLSTR LST generally had lower accuracy. The SWAs identified and described in this study may serve as alternative algorithms for retrieving LST products from SLSTR data.     

Retrieval of land surface temperature (LST) from landsat TM6 and TIRS data by single channel radiative transfer algorithm using satellite and ground-based inputs

The present study proposes land surface temperature (LST) retrieval from satellite-based thermal IR data by single channel radiative transfer algorithm using atmospheric correction parameters derived from satellite-based and in-situ data and land surface emissivity (LSE) derived by a hybrid LSE model. For example, atmospheric transmittance (τ) was derived from Terra MODIS spectral radiance in atmospheric window and absorption bands, whereas the atmospheric path radiance and sky radiance were estimated using satellite- and ground-based in-situ solar radiation, geographic location and observation conditions. The hybrid LSE model which is coupled with ground-based emissivity measurements is more versatile than the previous LSE models and yields improved emissivity values by knowledge-based approach. It uses NDVI-based and NDVI Threshold method (NDVI^THM) based algorithms and field-measured emissivity values. The model is applicable for dense vegetation cover, mixed vegetation cover, bare earth including coal mining related land surface classes. The study was conducted in a coalfield of India badly affected by coal fire for decades. In a coal fire affected coalfield, LST would provide precise temperature difference between thermally anomalous coal fire pixels and background pixels to facilitate coal fire detection and monitoring. The derived LST products of the present study were compared with radiant temperature images across some of the prominent coal fire locations in the study area by graphical means and by some standard mathematical dispersion coefficients such as coefficient of variation, coefficient of quartile deviation, coefficient of quartile deviation for 3rd quartile vs. maximum temperature, coefficient of mean deviation (about median) indicating significant increase in the temperature difference among the pixels. The average temperature slope between adjacent pixels, which increases the potential of coal fire pixel detection from background pixels, is significantly larger in the derived LST products than the corresponding radiant temperature images.     

Effects of prediction accuracy of the proportion of vegetation cover on land surface emissivity and temperature using the NDVI threshold method

Predicting land surface energy budgets requires precise information of land surface emissivity (LSE) and land surface temperature (LST). LST is one of the essential climate variables as well as an important parameter in the physics of land surface processes at local and global scales, while LSE is an indicator of the material composition. Despite the fact that there are numerous publications on methods and algorithms for computing LST and LSE using remotely sensed data, accurate prediction of these variables is still a challenging task. Among the existing approaches for calculating LSE and LST, particular attention has been paid to the normalised difference vegetation index threshold method (NDVI^THM), especially for agriculture and forest ecosystems. To apply NDVI^THM, knowledge of the proportion of vegetation cover (P_V) is essential. The objective of this study is to investigate the effect of the prediction accuracy of the P_V on the estimation of LSE and LST when using NDVI^THM. In August 2015, a field campaign was carried out in mixed temperate forest of the Bavarian Forest National Park, in southeastern Germany, coinciding with a Landsat-8 overpass. The P_V was measured in the field for 37 plots. Four different vegetation indices, as well as artificial neural network approaches, were used to estimate P_V and to compute LSE and LST. The results showed that the prediction accuracy of P_V improved using an artificial neural network (R²_CV = 0.64, RMSE_CV = 0.05) over classic vegetation indices (R²_CV = 0.42, RMSE_CV = 0.06). The results of this study also revealed that variation in the accuracy of the estimated P_V affected calculation results of the LSE. In addition, our findings revealed that, though LST depends on LSE, other parameters should also be taken into account when predicting LST, as more accurate LSE results did not increase the prediction accuracy of LST.     

A case study for intercomparison of land surface temperature retrieved from GOES and MODIS

In recent years, algorithms have been developed to derive land surface temperature (LST) from geostationary and polar satellite systems. However, few works have addressed the intercomparison between Geostationary Operational Environmental Satellites (GOES) and the available suite of polar sensors. In this study, differences in LSTs between GOES and MODerate resolution Imaging Spectroradiometer (MODIS) have been compared and also evaluated against ground observations. Due to the lack of split-window (SW) channels in the GOES M (12)-Q era, a dual-window algorithm using a mid-infrared 3.9 µm channel is compared with traditional SW algorithm. It is found that the differences in LST between different platforms are bigger during daytime than those during nighttime. During daytime, LSTs from GOES with the dual-window algorithm are warmer than MODIS LSTs, while LSTs from the SW algorithm are close to MODIS LSTs. The difference during daytime is found to be related to anisotropy in satellite viewing geometry, and land surface properties, such as vegetation cover and especially surface emissivity at middle infrared (MIR) channel. When evaluated against ground observations, the standard deviation (precision) error (2.35 K) from the dual window algorithm is worse than that (1.83 K) from the SW algorithm, indicating the lack of split-window channel in the GOES M(12)-Q era may degrade the performance of LST retrievals.     

Retrieval of land surface temperature from the Kalpana-1 VHRR data using a single-channel algorithm and its validation over western India

Indian geostationary satellite Kalpana-1 (K1) offers a potential to capture the diurnal cycle of land surface temperature (LST) through thermal infrared channel (10.5–12.5 μm) observations of the Very High Resolution Radiometer (VHRR) sensor. A study was carried out to retrieve LST by adapting a generalized single-channel (SC) algorithm (Jiménez-Muñoz and Sobrino, 2003) for the VHRR sensor over India. The basis of SC algorithm depends on the concept of Atmospheric Functions (AFs) that are dependent on transmissivity, upwelling and downwelling radiances of the atmosphere. In the present study AFs were computed for the VHRR sensor through the MODTRAN simulations based upon varying atmospheric and surface inputs. The AFs were fitted with the atmospheric columnar water vapour content and a set of coefficients was derived for LST retrieval. The K1-LST derived with the SC algorithm was validated with (a) in situ measurements at two sites located in western parts of India and (b) the MODIS LST products. Comparison of K1-LST with the in situ measurements demonstrated that SC algorithm was successful in capturing the prominent diurnal variations of 283–332 K in the LST at desert and agriculture experimental sites with a rmse of 1.6 K and 2.7 K, respectively. Inter comparison of K1-LST and MODIS LST showed a reasonable agreement between these two retrievals up to LST of 300 K, however a cold bias up to 7.9 K was observed in MODIS LST for higher LST values (310–330 K) over the hot desert region.     

热红外遥感地表温度与发射率地面验证进展

地表温度与发射率是地表—大气系统长波辐射和潜热通量交换的直接驱动力,是描述区域和全球尺度上地表能量平衡与水平衡的重要参数,其时空变化信息在气象预测、气候变化、水循环、地质勘探、农林监测和城市热环境等诸多领域具有广泛的应用.热红外遥感作为当前获取区域或全球尺度上地表温度和发射率的最有效手段之一,相较于传统的地面点位测量方...     

陆地卫星TM6波段范围内地表比辐射率的估计

地表比辐射率是用热红外波段遥感数据反演地表温度的关键参数。目前,应用陆地卫星TM6波段数据反演地表温度共有3种算法,即大气校正法、单窗算法和单通道算法。这3种算法都需要TM6波段范围内的地表比辐射率作为地表温度反演参数。本文首先简介这3种反演算法;然后着重探讨TM6波段地表比辐射率估计方法;最后,利用这一方法对山东省陵县附近农田地区进行地表比辐射率估计和地表温度反演。结果表明,该方法能获得较合理的地表温度反演结果。     

中国陆地区域陆表温度业务化遥感反演算法及产品运行系统

地表温度反演的裂窗算法已成功应用于NOAA系列卫星热红外遥感数据。目前，裂窗算法中应用较为广泛的一种是Becker等人于1990年提出的局地裂窗算法，主要是通过辐射传输模型模拟不同地表条件和大气状况下，地表温度和发射率对红外辐射亮温的影响，从而发展出一个利用AVHRR4，5通道亮温数据反演地表温度的线性模型。在晴空无云和地表比辐射率能精确估算的情况下，Becker算法反演地表温度的精度在1K以内。Becker算法用Lowtran程序模拟计算地表辐射量，且模型中参数主要针对NOAA-9传感器特性得到。本文在Becker算法的基础上，针对NOAA-16／17传感器热红外通道光谱响应函数特性，利用最新的、计算光谱分辨率更高的MODTRAN程序模拟不同大气状况下，不同地表温度和发射率对NOAAAVHRR4，5通道辐射亮温响应特性的影响，改进Becker算法中模型参数，使之能适用于NOAA-16／17热红外数据。同时，本文利用植被指数NDVI，在中国陆地区域lkm分辨率最新地表分类数据的基础上，得到模型中需要的地表比辐射率参数，将改进的模型应用于1km分辨率NOAA17数据，得到了旬合成中国陆地区域范围地表温度，通过地面气象台站实测数据对比验证．取得了较好的结果。     

针对Terra/MODIS数据的改进分裂窗地表温度反演算法

针对Terra/MODIS数据提出改进的分裂窗地表温度反演算法。充分考虑了传感器观测角度(VZA)的影响,并对地表和有效大气辐射按照不同的亮度温度区间分别进行Planck函数简化。利用TIGR3大气廓线库中的875条晴空大气廓线,ASTER波谱库中的106条地物发射率波谱,结合MODTRAN4大气辐射传输模型模拟得到分裂窗算法系数。利用MODTRAN4模拟数据对算法精度进行验证,结果表明本文的改进算法和原算法的均方根误差RMSE分别为0.34K和0.65K。敏感性分析表明,在中等湿润的大气条件下,算法对大气水汽含量并不敏感。该算法降低了传感器观测角度带来的地表温度反演误差。利用2009年6月美国SURFRAD辐射观测网6个站点的实测数据对改进算法、原算法以及MOD11_L2地表温度产品进行了对比验证,RMSE分别是0.93K、1.49K和1.0K,表明本文算法可以提高反演精度。     

Inverse problems theory and application: analysis of the two-temperature method for land-surface temperature and emissivity estimation

The two-temperature method (TTM) allows the separation of land-surface temperature and land-surface emissivity information from radiance measurements, and therefore, the solution can be uniquely determined by the data. However, the inverse problem is still an ill-posed problem, since the solution does not depend continuously on the data. Accordingly, we have used some mathematical tools, which are suited for analyses of ill-posed problems in order to show TTM properties, evaluate it, and optimize its estimations. Related to this last point, we have shown that it is necessary to constrain the problem, either by defining a region of physically admissible solutions and/or by using regularization methods, in order to obtain stable results. Besides, the results may be improved by using TTM with systems that possess a high temporal resolution, as well as by acquiring observations near the maximum and minimum of the diurnal temperature range.     

单窗算法和单通道算法对参数估计误差的敏感性分析

单窗算法和单通道算法的提出为应用LandsatTM热波段反演地表温度开辟了新途径。由于进行地面像元尺度同步温度测量难度较大,目前尚无法对这两种算法进行直接评判。通过对算法所需参数分别取一定变动区间并进行渐变取值测算,分析了算法对其参数估计误差的敏感性,以此作为对这两种算法适用性的一个间接评判依据。     

耦合双时相与邻近像元信息的极轨卫星地表组分温度反演

与混合像元的地表温度相比,植被和土壤的组分温度具有更明确的物理意义.因此,本文提出了一种从具有广泛应用的极轨卫星地表温度产品中分离出植被和土壤组分温度的算法.该算法使用温度日变化模型作为桥梁连接极轨卫星一日内的两次观测,利用多像元数据进行模型求解,从而得到过境时刻的地表植被和土壤组分温度.论文针对MODIS数据开展了地...     

Estimating land-surface temperature under clouds using MSG/SEVIRI observations

The retrieval of land-surface temperature (LST) from thermal infrared satellite sensor observations is known to suffer from cloud contamination. Hence few studies focus on LST retrieval under cloudy conditions. In this paper a temporal neighboring-pixel approach is presented that reconstructs the diurnal cycle of LST by exploiting the temporal domain offered by geo-stationary satellite observations (i.e. MSG/SEVIRI), and yields LST estimates even for overcast moments when satellite sensor can only record cloud-top temperatures. Contrasting to the neighboring pixel approach as presented by Jin and Dickinson (2002), our approach naturally satisfies all sorts of spatial homogeneity assumptions and is hence more suited for earth surfaces characterized by scattered land-use practices. Validation is performed against in situ measurements of infrared land-surface temperature obtained at two validation sites in Africa. Results vary and show a bias of −3.68 K and a RMSE of 5.55 K for the validation site in Kenya, while results obtained over the site in Burkina Faso are more encouraging with a bias of 0.37 K and RMSE of 5.11 K. Error analysis reveals that uncertainty of the estimation of cloudy sky LST is attributed to errors in estimation of the underlying clear sky LST, all-sky global radiation, and inaccuracies inherent to the ‘neighboring pixel’ scheme itself. An error propagation model applied for the proposed temporal neighboring-pixel approach reveals that the absolute error of the obtained cloudy sky LST is less than 1.5 K in the best case scenario, and the uncertainty increases linearly with the absolute error of clear sky LST. Despite this uncertainty, the proposed method is practical for retrieving the LST under a cloudy sky condition, and it is promising to reconstruct diurnal LST cycles from geo-stationary satellite observations.     

Land surface temperature retrieval and its validation using NOAA AVHRR thermal data

The retrieval of land (soil-vegetation complex) surface temperature (LST) was carried out over semi-arid mixed agriculture landscape of Gujarat using thermal bands (channel 4 and 5) and ground emissivity from atmospherically corrected NDVI of NOAA AVHRR LAC images. The atmospheric correction of Visible and NIR band reflectance was done using SMAC model. The LST computed from split-window method and subsequently corrected with fractional vegetation cover were then compared with near synchronous ground observations of soil and air temperatures made during 13–17 January and April, 1997 at five Land Surface Processes Experiment (LASPEX) sites of Anand, Sanand, Derol, Arnej and Khandha covering 100 km x 100 km. The fractional vegetation cover corrected LST at noon hrs. varied from 301.6 – 311.9K in January and from 315.8 – 325.6K in April. The LST_corr were found to lie in the mid way between AT and ST during January. But in April, LST were found to be more close to ST which may be due to relatively poor vegetation growth as indicated by lower NDVI values in April indicating more contribution to LST from exposed soil surface.     

A radiance-based split-window algorithm for land surface temperature retrieval: Theory and application to MODIS data

The split-window algorithm is the most commonly used method for land surface temperature (LST) retrieval from satellite data. Simplification of the Planck’s function, as an important step in developing the SWA, allows us to directly relate the radiance to the temperature toward solving the radiative transfer equation (RTE) set. In this study, Planck’s radiance relationship between two adjacent thermal infrared channels was modeled to solve the RTE set instead of simplification of the Planck’s function. A radiance-based split-window algorithm (RBSWA) was developed and applied to Moderate Resolution Imaging Spectroradiometer (MODIS) data. The performance of the RBSWA was assessed and compared with three most common brightness temperature-based split-window algorithms (BTBSWAs) by using the simulated data and satellite measurements. Simulation analysis showed that the LST retrieval using RBSWA had a Root Mean Square Error (RMSE) of 0.5 K and achieved an improvement of 0.3 K compared with three BTBSWAs, and the LST retrieval accuracy using RBSWA was better than 1.5 K considering uncertainties in input parameters based on the sensitivity analysis. For application of RBSWA to MODIS data, the results showed that: 1) comparison between LST from MODIS LST product and LST retrieved using RBSWA showed a mean RMSE of 1.33 K for 108 groups of MODIS image covering continental US, which indicates RBSWA is reliable and robust; 2) when using the measurements from US surface radiation budget network as real values the RMSE of the RBSWA algorithm was 2.55 K and was slightly better than MODIS LST product; and 3) through the cross validation using Advanced Spaceborne Thermal Emission and Reflection Radiometer LST product, the RMSE of the RBSWA algorithm was 2.23 K and was 0.28 K less than that of MODIS LST product. We conclude that the RBSWA for LST retrieval from MODIS data can attain a better accuracy than the BTBSWA.     

藏东南冰川地区250 m空间分辨率全天候地表温度生成

遥感全天候地表温度产品在多云雾地区意义重大,对冰川泥石流多发的藏东南地区极具应用价值,但遥感全天候地表温度空间分辨率不足限制了其在精细化灾害监测中的应用。以藏东南冰川地区为研究区,采用高程、坡度、坡向、地表覆盖类型、植被指数、地表反射率、积雪指数作为全天候地表温度的影响因子,结合移动窗口,进行多种地表温度降尺度方法的对比,进而使用最优的降尺度方法将现有的遥感全天候地表温度产品（TRIMS LST）的空间分辨率从1 km提升至250 m。利用地面站点实测数据的评价结果表明,基于梯度提升决策树（LightGBM）的降尺度方法得到的250 m空间分辨率全天候地表温度的均方根误差在白天/夜间为2.25 K/2.15 K,优于基于多元线性回归和随机森林的降尺度方法,且比原始1 km分辨率全天候地表温度的精度高0.25 K左右。基于Q指数与SIFI指数的图像质量评价结果表明,降尺度得到的250 m地表温度不仅在空间格局和幅值上与原始1 km遥感全天候地表温度一致,而且补充了大量的地表温度空间细节信息。生成得到的250 m分辨率的地表温度对于藏东南冰川地区的灾害分析具有积极的意义。     

Detection of geothermal anomalies using Landsat 8 TIRS data in Tulu Moye geothermal prospect,Main Ethiopian Rift

Despite the high geothermal potential of the Main Ethiopian Rift (MER), risks associated with the industry and the difficulty of identifying possible targets using ground surveys alone continue to impede the development of geothermal power diligence in Ethiopia. In this paper, we investigate the geothermal potential of the Tulu Moye prospect area in the MER using Landsat 8, which is an important and cost-effective method of detecting geothermal anomalies. Data with a path/row of 168/054 were obtained from the Landsat 8 Operational Land Imager (OLI) and Thermal Infrared (TIR) sensors for October 17, 2014. Based on radiometric calibration, atmospheric correction (with the 6S model) and an NDVI-based threshold method for calculating land surface emissivity, a split-window algorithm was applied to retrieve the land surface temperature (LST) of the study area. Results show LST values ranging from 292.2 to 315.8 K, with the highest values found in barren lands. A comparison of LST between the Moderate Resolution Imaging Spectroradiometer (MODIS) and Landsat 8 shows a maximum difference of 1.47 K. Anomalous areas were also discovered, where LST was about 3-9 K higher than the background area. We identified seven of these as areas of high geothermal activity in the Tulu Moye prospective geothermal area. Auxiliary data and overlay analysis tools eliminated any non-geothermal influences. The research reveals that the distribution of highy prospective geothermal areas is consistent with the development and distribution of faults in the study area. Magmatism is the thermal source and faults provide conduits for the heat to flow from earth’s interior to the surface, facilitating the presence of geothermal anomalies. Finally, TIR remote sensing methods prove to be a robust and cost-effective technique for detecting LST anomalies in the geologically active area of MER. Moreover, combining TIR remote sensing with knowledge of the structural geology of the area and geothermal mechanisms is an efficient approach to detecting geothermal areas.     

地表温度日变化模型偏差系数解算的地表温度降尺度

地表温度LST(Land Surface Temperature)是全球气候变化研究的关键参数,遥感是获取全球和区域尺度地表温度的一种切实可行手段,但现有的单一传感器无法提供高时空分辨率的LST数据,限制了遥感地表温度数据的深入广泛应用.现有的降尺度方法难以生成无缝高时空分辨率的地表温度数据,且降尺度效果易受高空间分辨...     

基于城市地表温度与归一化植被指数对北京市建成区热环境分析

针对研究城市热环境的过程中,利用归一化植被指数(NDVI)进行地表温度(LST)反演,再将LST和NDVI结合说明地物变迁与城市热环境的影响的现状,利用Landsat-8多时相遥感影像、高分辨率影像、公开GIS等多源数据,通过人工交互判读和量化统计分析,实现了2013—2017年北京建成区NDVI变化及其对地表热环境影响分析,再对分析结果进行差值拟合评价。对NDVI阈值分割按照大小为LC1、LC2、LC3、LC4,对LST分为高温区(TH)、常温区(TR)和低温区(TL)。结果表明,2013年至2017年:1)建成区的平均NDVI增加0.03,其中LC1增加1.0%,LC2减少11.6%,LC3区域减少1.7%,LC4区域增加12.3%。2)建成区平均LST增加2.55 K,TH百分比增加0.6%;TR百分比减少1.1%,TL百分比增加0.5%。3)NDVI相对增加,地表温度相对下降以及NDVI相对减少,地表温度相对上升占60%,NDVI相对下降,地表温度相对下降以及NDVI相对增加,地表温度相对上升的占40%。     

庐山地区地表温度反演分析

研究庐山地区的地表温度,对于本地区的生态保护和开发具有重要意义。本文使用了目标区域内Landsat TM/ETM+从2000~2010年间共8幅影像数据,利用单通道算法和NDVI地表辐射率估算方法反演了区域内的温度场,并生成了历年温度对比效果图。分析推断目标区域内的温度场可能存在一个短年限的温度变化起伏规律,对庐山周边山区的影响有待进一步研究。