Abstract: Land surface temperature is one of the most important factors affecting urban environmental quality. In the past few years, the study of land surface temperature is drawing more and more attention in major cities. In this paper, based on the image data of the Landsat series satellites , using RSTAR radiation transfer model to retrieve land surface temperature in Sanya, the land surface temperature products in 2000, 2004, 2008, 2012, 2016 and 2018 were obtained. The research results show that the method is easy to operate and has improven the inversion accuracy significantly. It can be widely used in land surface temperature inversion of other cities, thereby of high value of application.
Keywords: land surface temperature; radiative transfer; Landsat; Sanya
|Title||Inverted land surface temperature for Sanya city based on Landsat data|
|Data corresponding author||Hu Die（email@example.com）|
|Data authors||Gu Yanchun, Hu Die, Zhang Ying, Zhang Linlin|
|Geographical scope||Sanya (18°09′34″–18°37′27″N, 108°56′30″–109°48′28″E)|
|Spatial resolution||Landsat 5: 30/120 m|
Landsat 8: 30/100 m
|Data volume||427 KB|
|Data format||*.vsd, *.png, *.pro|
|Data service system||<http://www.sciencedb.cn/dataSet/handle/675>|
|Sources of funding||Major Science and Technology Program of Hainan Province (ZDKJ2016021); Major Science and Technology Program of Hainan Province (ZDKJ2017009); Science and Technology Program of Sichuan Province (2018JZ0054).|
|Dataset composition||This dataset mainly includes the following four folders:1. cover folder: which contains a map of administrative zoning in Sanya in PNG format; 2. Sanya temperature inversion map folder: which contains the temperature inversion images of Sanya during 2000-2018 in PNG format; 3. data set construction and processing process: which contains land surface temperature inversion data based on the radiative transfer model in VSD format; 4. Temperature inversion code folder: which contains Landsat 5(LST5) and Landsat 8 (LST8) temperature inversion code data in PRO format.|
Land surface temperature (LST) is an important input parameter for land surface research analysis and application model creation in environmental monitoring, climate, hydrology, agriculture, etc. It is also the basis for many other research projects and applications, such as monitoring soil moisture, enhancing position discrimination of geothermal energy, etc. are all inseparable from the determination of land surface temperature . Inversion of land surface temperature has always been an important role of remote sensing quantitative application. Choosing the best land surface temperature inversion algorithm in specific research is the primary factor to improve the accuracy of inversion.
At present, many different algorithms have been proposed for land surface temperature inversion, such as image-based inversion algorithm , single-channel algorithm[3,4] , single-window algorithm[5,6,7], and window algorithm[8,9,10,11,12,13] and multi-channel algorithms[14,15,16] . The most commonly used is the windowing algorithm proposed by Becker et al. in 1990. This algorithm has been widely used in NOAA series satellite thermal infrared remote sensing data. However, due to the heterogeneity, complexity and atmospheric effects of terrestrial surface materials, the inversion of land surface temperature by remote sensing images has always been an important and complex research topic. The RSTAR radiation transmission model fully considers these problems and makes up for the shortcomings of long-sequence of land surface temperature inversion, such as the atmospheric parameters are not considered; the physical meaning of atmospheric transmission is not clear;it has a high accuracy only in the clear sky without clouds and accurate estimation of surface emissivity; the NOAA series satellite thermal infrared remote sensing data is the best; it is difficult to determine the general radiation transfer equation method parameters, and the temperature inversion process is complicated. As the number of remote sensing satellites increases year by year, remote sensing images become more and more abundant. Therefore, the remote sensing can be used for temperature inversion of long-term sequences, which can realize the advantages of regional temperature monitoring.
In this paper, the Landsat 5 and Landsat 8 image data are used to calculate the atmospheric profile based on the RSTAR radiation transmission model. The simple and accurate surface emissivity estimation method proposed by Zhai Zhihaois used to calculate the Sanya City in 2000 and 2004. In 2008, 2012, 2016 and the summer of 2018 (June-August), there are 6 scenes (no cloud or less cloud) land surface temperature, so as to monitor the trend of land surface temperature changes in Sanya.
The research area of Sanya (18°09′34′′–18°37′27′′N, 108°56′30′′–109°48′28′′E) is located at the southernmost tip of Hainan Island, adjacent to Lingshui County in the east. It is connected to Ledong County in the west, Baoting County in the north and Nanhai in the south. The climate is a tropical marine monsoon climate with an average annual temperature of 25.7 °C.
The data set construction process includes data downloading, calculation of atmospheric up-and-down radiation and atmospheric transmittance parameters, temperature inversion code implementation, cropping and output steps, as shown in Figure 1.
The IDL used in the data processing process is a data analysis generated by the imaging application program and programming language, which has the advantages of flexible data input and output, rapid visualization, and support for external language interfaces. The best choice, this data set uses the IDL 8.5 version.
2.1 Data source and data summary
The Landsat series of image data has the advantage of relatively high spatial resolution (such as the spatial resolution of the Landsat 8 thermal infrared band of 100 m), which helps to improve the accuracy of the land surface temperature inversion results. On March 1, 1984, the USGS successfully launched the Landsat 5 satellite with a thermal infrared band Band 6 (10.40 -12.5μm); on February 11, 2013, NASA successfully launched the Landsat 8 satellite with two thermal infrared bands Band 10 (10.6-11.2μm) and Band 11 (11.2-12.5μm), due to problems with Band 11 operation, this paper uses Band 10 single-band to invert the surface temperature. The Landsat series of image data used in this article can be obtained from the Geospatial Data Cloud website (http://www.gscloud.cn/search).
2.2 Data processing
2.2.1 Atmospheric profile parameter calculation
The calculation of atmospheric profile parameters is mainly based on the atmospheric radiation transmission mode to calculate the atmospheric up-radiation, down-radiation and atmospheric transmittance parameters, and the atmospheric transmission mode is the RSTAR radiation transmission mode. This paper calculates the image of each image directly from the atmospheric correction parameter calculator website (https://atmcorr.gsfc.nasa.gov/) according to the transit time, central latitude and longitude, quarter (summer or winter) and sensor type of the specific image. Atmospheric up-radiation, down-radiation, and atmospheric transmittance parameters.
2.2.2 Visible band processing
According to Chander et al., when comparing images from different sensors, the Top of Atmosphere Reflectance (TOA reflectance) should be used as follows: First, eliminate the error caused by the cosine of different solar zenith angles; Secondly, TOA reflectance is used to compensate for the different values of solar irradiance caused by spectral band difference. Finally, TOA reflectance is used to correct the change of solar-terrestrial distance caused by different data acquisition time. The key code for this process (Figure 2) is done through code in IDL 8.5.
2.2.3 Surface emissivity calculation
Based on the inversion algorithm of surface emissivity proposed by Qin Zhihao, after the water body is masked, the surface is divided into two categories: natural (vegetation + bare soil) and town (vegetation + building surface). The algorithm has three main steps: first, calculate the vegetation coverage; secondly, the urban surface can be regarded as composed of different proportions of vegetation and buildings, and the natural surface can be regarded as composed of different proportions of vegetation and bare soil; finally, generated the surface emissivity map of the entire study area image. The key code of the process is shown in Figure 3.
2.2.4 land surface temperature retrieval
This section is mainly divided into three steps: First, the atmospheric up-radiation, down-radiation and atmospheric transmittance parameters calculated from the atmospheric profile parameters are input into the code shown in Figure 4; secondly, obtain the inversion temperature map based on the data calculated by visible light band processing and surface emissivity; finally, select the storage location for the temperature inversion result. The key code of the process is shown in Figure 4.
Selecting the calculation method of radiation transfer equation in various land surface temperature inversion algorithms., the atmospheric profile is calculated based on the RSTAR radiation transfer model. The surface emissivity estimation method proposed by Qin Zhihaois used to calculate 2000, 2004 and 2008. In 2012, 2016 and 2018, there are 6 scenes (no cloud or less cloud) Sanya city land surface temperature, some results are shown in Figure 5.
This data set is quality controlled through the following aspects:
(1) The quality control of data source. When selecting Landsat satellite data, there are data, no cloud or less clouds, no strips as the selection criteria. Because high quality data sources are necessary for high precision results.
(2) The quality control of temperature inversion results. In this paper, in order to verify the accuracy of land surface temperature products, based on the correlation between the temperature obtained from the statistical analysis of Sanya city temperature data and the published simultaneous surface temperature, the correlation coefficient R2 is 0.91, and this is used as the verification of land surface temperature products. A reliable basis for data.
This study puts forward a surface temperature inversion technique suitable for urban thermal environment characteristics, and provides technical support for remote sensing services in urban construction. Data results are highly practical and universal. For example, inversion products can provide data product services for urban ecological environment research, regardless of region size. At the same time, the data results can provide a detection method for the environmental protection department and the urban construction department to carry out urban thermal environment change monitoring, which has strong application value and practicability.
Inversion of urban thermal space information parameters by means of remote sensing, giving full play to the periodicity and timeliness of remote sensing monitoring, and facilitating rapid analysis and judgment of spatial and temporal evolution characteristics of urban ecological environment in a short time.
 Chen Yuanyuan. Research on surface temperature inversion algorithm of high score No.5 thermal infrared data[D]. Beijing: Chinese Academy of Agricultural Sciences, 2017.
 Wan Z, Dozier J. A generalized split-window algorithm for retrieving land-surface temperature from space[J]. IEEE Transactions on Geoscience & Remote Sensing, 1996, 34(4): 892-905.
 Sobrino J A, Jimenez-Munoz J C, Soria G, et al. Land Surface Emissivity Retrieval From Different VNIR and TIR Sensors[J]. IEEE Transactions on Geoscience & Remote Sensing, 2008, 46(2): 316-327.
 Nichol J E. High-resolution surface temperature patterns related to urban morphology in a tropical city: A satellite-based stud[J]. Journal of Applied Meteorology, 1996, 35(1): 135-46.
 Zhou Ji, Li Jing, Zhao Xiang, et al. Modified Single Channel Algorithm for Retrieving Surface Temperature Using HJ-1B Satellite Data[J]. Journal of Infrared and Millimeter Waves, 2011, 30(1): 61-67.
 Hao Q Z , Hua Z M , Karnieli A , et al. Mono-window Algorithm for Retrieving Land Surface Temperature from Landsat TM6 data[J]. Acta Geographica Sinica, 2001, 56(4):456-466.
 Jiménez-Muoz, Juan C . A generalized single-channel method for retrieving land surface temperature from remote sensing data[J]. Journal of Geophysical Research, 2003, 108(D22): 4688.
 Becker F, Li Z L. Towards a local split window method over land surfaces. International Journal of Remote Sensing[J]. International Journal of Remote Sensing, 1990, 11(3): 369-393.
 Wan Z. New refinements and validation of the MODIS Land-Surface Temperature/Emissivity products[J]. Remote Sensing of Environment, 2014, 140(1): 36-45.
 Jimenez-Munoz J C, Sobrino J A. Split-Window Coefficients for Land Surface Temperature Retrieval From Low-Resolution Thermal Infrared Sensors[J]. IEEE Geoscience & Remote Sensing Letters, 2008, 5(4): 806-809.
 Wigneron J P, Schmugge T, Chanzy A, et al. Use of passive microwave remote sensing to monitor soil moisture[J]. Agronomie, 1998, 18(1): 27-43.
 Rozenstein O, Qin Z, Derimian Y, et al. Derivation of land surface temperature for Landsat-8 TIRS using a split window algorithm[J]. Sensors, 2014, 14(4): 5768-5780.
 Wang J, Huang B, Fu D, et al. Spatiotemporal Variation in Surface Urban Heat Island Intensity and Associated Determinants across Major Chinese Cities[J]. Remote Sensing, 2015, 7(4): 3670-3689.
 Peng S, Piao S, Ciais P, et al. Surface urban heat island across 419 global big cities.[J]. Environmental Science & Technology, 2012, 46(2): 696-703.
 Henry J A, Dicks S E, Wetterqvist O F, et al. Comparison of satellite, ground-based, and modeling techniques for analyzing the urban heat island[J]. Photogrammetric Engineering & Remote Sensing, 1989, 55(1): 69-76.
 Imhoff M L, Zhang P, Wolfe R E, et al. Remote sensing of the urban heat island effect across biomes in the continental USA[J]. Remote Sensing of Environment, 2010, 114(3): 504-513.
 Yang Hu, Yang Zhongdong. Remote Sensing Inversion Algorithm and Product Operation System for Land Surface Temperature Operation in China[J]. Journal of Remote Sensing, 2006, 10(4): 600-607.
 Qin Zhihao, Zhang M H, Arnon K, et al. Single window algorithm for calculating surface temperature using land satellite TM6 data[J]. Journal of Geographical Sciences, 2001, 56(4): 456-466.
Gu Yanchun, Hu Die, Zhang Ying, Zhang Linlin. Inverted land surface temperature for Sanya city based on Landsat data[DB/OL]. Science Data Bank, 2018. (2019-02-22). DOI: 10.11922/sciencedb.675.
How to cite this article
Gu Yanchun, Hu Die, Zhang Ying, Zhang Linlin. Inverted land surface temperature for Sanya city based on Landsat data[J/OL]. China Scientific Data, 2019, 4(2). (2019-04-19). DOI: 10.11922/csdata.2018.0070.zh.