RICE-HDM: a Remote Sensing Framework for Monitoring Rice High-Temperature Heat Damage in the Middle and Lower Reaches of the Yangtze River

Li Liu,Cian Yuan,Yan Chen,Zixuan Chen,Dong Ren,Biao Xiong
DOI: https://doi.org/10.1145/3653644.3665208
2024-01-01
Abstract:In the middle and lower reaches of the Yangtze River Basin in China, a large amount of rice enters the heading and flowering stage during the hot summer following the plum rain season. The prolonged high temperature easily triggers widespread rice high-temperature heat damage, resulting in severe economic losses. This paper proposes a based remote sensing paddy rice high-temperature heat damage monitoring (RICE-HDM) framework, to monitor the high-temperature heat damage in Hubei, Hunan, Jiangxi, Jiangsu, and Anhui province, from 2003 to 2019. Firstly, based on the Moderate Resolution Imaging Spectroradiometer (MODIS) Land Surface Temperature (LST) products and other data, we use Random Forest model to estimate the daily average temperatures in the study area. Secondly, a multi-platform data fusion and spatial interpolation module for processing daily average temperature images is proposed to construct a fully covered high-precision remote sensing dataset of daily average temperature. Finally, we utilize the produced daily average temperature dataset and rice information dataset obtained from our previous studies to estimate the proportion of rice heat damage areas from 2003 to 2019 of study area and to conduct disaster extent assessments, producing disaster distribution maps. The experimental results show that the proposed framework achieves quantitative and accurate monitoring of rice high-temperature heat damage, which is helpful for guiding real-time monitoring of the disaster and the assessment of the disaster.
What problem does this paper attempt to address?