The Innovation Team of Agricultural Remote Sensing of the Institute of Agricultural Resources and Regional Planning (IARRP) of the Chinese Academy of Agricultural Sciences (CAAS), has made significant progress in the temporal normalization of polar-orbiting satellite-derived land surface temperature. The research results, titled "Development of a hybrid algorithm for temporal normalization of polar-orbiting satellite-derived land surface temperature" and "An Improved Integrated Model for Temporal Normalization of Satellite-derived Land Surface Temperature," were published in the flagship journal of the remote sensing field, "IEEE Transactions on Geoscience and Remote Sensing."

Land surface temperature is a crucial parameter in fields such as agriculture, meteorology, ecology, and hydrology. However, due to the scanning characteristics of remote sensing instruments, the inconsistent observation times of polar-orbiting satellites have limited the widespread application of land surface temperature data. Despite the development of various algorithms to address this issue, an effective solution for the non-linear changes in temperature during clear-sky conditions in the daytime period has been lacking.

To overcome the limitations of previous algorithms, which could only handle clear-sky conditions for the entire daytime period or linear changes during the morning, the agricultural remote sensing team developed two types of land surface temperature temporal normalization models based on the time-varying characteristics of land surface temperature during the overpass periods of polar-orbiting satellites. This was done to more accurately parameterize the modeling of land surface temperature over time. Additionally, the team members analyzed the influence of various factors such as different land surface characteristics, atmosphere, and solar radiation, and introduced an integrated regression algorithm to optimize the estimation of model parameters. By combining actual observations, remote sensing modeling, and data mining techniques, the research integrated the advantages of multiple methods, thereby enhancing the applicability and accuracy of the land surface temperature temporal normalization algorithm. The research results provide important theoretical and methodological support for the monitoring and management of agricultural ecological environments at the regional scale.

Figure 1. Cross-validation of MODIS land surface temperature before and after normalization with time-consistent GOES geostationary satellite land surface temperature

Dr. Du Wenhui from the IARRP is the first author of this paper. This research was supported by the National Natural Science Foundation of China's Innovative Research Group Project "Mechanism and Methods of Agricultural Remote Sensing."

Paper Links:

1. https://ieeexplore.ieee.org/document/10214594 

2. https://ieeexplore.ieee.org/document/10457853