| All-weather surface temperature products are widely used in research on climate change and disaster prevention and control because they are not affected by cloud cover and long-term sequence.They have high application potential and value in cloudy and foggy areas such as the glacial area of the Qinghai-Tibet Plateau.However,at present,the spatial resolution of all-weather surface temperature products of remote sensing is mostly in 1 km.In more refined disaster monitoring such as glacial debris flow,the spatial resolution is insufficient,which affects the further application of surface temperature.In order to improve the spatial resolution of remote sensing all-weather surface temperature products in the glacial area of the Qinghai-Tibet Plateau,the main research work carried out in this paper is as follows:(1)A 250 m all-weather surface temperature generation method for the glacial area of the Qinghai-Tibet Plateau is proposed.In this paper,by analyzing the relationship between all-weather land surface temperature and its temporal and spatial influencing factors,elevation,land cover type,vegetation index,snow cover index,band reflectivity,and other data,a downscaling model of all-weather land surface temperature is constructed,and the space of all-weather land surface temperature products The resolution has been increased from 1 km to 250 m.Compared the accuracy performance of a variety of linear and machine learning regression algorithms used in the model.The comparison results show that the Light GBM machine learning method has the highest accuracy.Verification by the measured data from the ground station shows that the RMSE of the downscaled surface temperature at the site during the day and night is about 2.25 K and 2.16 K,respectively,which is about 0.5 K higher than the accuracy of the original1 km surface temperature product.The calculation results of the image quality index show that the downscaled surface temperature not only obtains a large amount of detailed thermal information but also maintains a high degree of consistency with the original 1km surface temperature in terms of spatial pattern and amplitude.(2)Based on the 250-meter downscale all-weather surface temperature product,a100-meter-level all-weather surface temperature generation method for the glacial area of the Qinghai-Tibet Plateau is proposed,and the application range of this method is extended to the entire Qinghai-Tibet Plateau.After obtaining the 250 m all-weather surface temperature product,the problem of greatly reducing the impact factor of the surface temperature is faced when the spatial resolution is further improved.In order to obtain a surface temperature product with higher spatial resolution,this paper analyzed the relationship between elevation,slope aspect,land cover type,and surface temperature at a spatial resolution of 250 m,and redesigned and constructed a high-resolution allweather surface temperature drop.The scale model increases the spatial resolution of the all-weather surface temperature product from 250 m to the 100-meter level.Among the multiple regression algorithms used in the model,the Light GBM machine learning method still has the highest accuracy.The site verification results show that the RMSE of the 100-meter downscaled surface temperature at the site is about 2.01 K and 2.04 K during the day and at night,respectively,maintaining good accuracy.The calculation results of the image quality index show that the 100-meter downscaled surface temperature product once again obtains a large amount of detailed thermal information on the basis of the 250 m surface temperature,and maintains a high degree of consistency in the spatial pattern and amplitude.In addition,this paper selected three other typical areas of the Qinghai-Tibet Plateau for downscaling analysis to verify whether the proposed downscaling method can be applied to other areas of the Qinghai-Tibet Plateau.The verification results show that the RMSE of the 250 m and 100-meter downscaled surface temperatures at the site are about 3.08 K and 2.91 K during the day,and 2.7 K and 2.69 K at night,respectively.Therefore,the downscaling method proposed in this paper can be effectively applied to the all-weather surface temperature downscaling process in most areas of the Qinghai-Tibet Plateau.(3)Time series analysis of the generated high-resolution all-weather surface temperature products and analysis of the relationship between the glacial debris flow disasters in the region and the high-resolution all-weather surface temperature.This paper uses the linear trend estimation method to analyze the high-resolution all-weather surface temperature temporal and spatial changes in the glacial area of the Qinghai-Tibet Plateau from 2003 to 2019.The results show that more than 80% of the area has been in a warming trend in the past 20 years,and the average warming trend during the day and night is 0.13 K and 0.06 K per year,respectively.The spatial distribution of surface temperature is mainly affected by topography,with low temperatures at the top of mountains and high temperatures at the bottom of ravines.This paper combines highresolution all-weather surface temperature products and precipitation data to study the occurrence conditions of glacial debris flow.The analysis results show that continuous high temperature,continuous precipitation,and sudden changes in surface temperature may cause glacier debris flow disasters.The high-resolution all-weather surface temperature generation method proposed in this paper in the glacial area of the Qinghai-Tibet Plateau maintains high accuracy and image quality while increasing the spatial resolution of all-weather surface temperature from 1 km to 100 meters,which is a refined glacial debris flow in this area.Researches on disaster prevention and control have provided reliable surface temperature data.The time series of high-resolution all-weather surface temperature in the glacial region of the Qinghai-Tibet Plateau and its influence on the glacial debris flow are discussed.The research results have a certain significance for the high-resolution all-weather surface temperature generation and disaster monitoring in the glacial region of the Qinghai-Tibet Plateau. |
PDF Full Text Request