Publications

Zhang, GX; Zhang, GX; Zhu, SY; Zhang, N; Xu, YM (2023). Estimation on the Hourly Distribution of Near-Surface Temperature Lapse Rate Under Winter Clear-Sky Conditions. IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 61, 4107013.

Abstract
The near-surface (2 m) temperature lapse rate (TLR) is a key parameter in various environmental studies. However, high spatial and temporal resolution TLRs are not usually available on regional scales, especially in mountainous regions. The purpose of this study is to model spatiotemporal continuous TLR in a mountain area using observed air temperature, Moderate Resolution Imaging Spectroradiometer (MODIS) land surface temperature (LST) products, land cover maps, and the ASTER digital elevation model (DEM). To address this issue, the sliding window method was employed in this study to model TLR from MODIS LST, and the diurnal temperature cycle (DTC) model was utilized to fit the diurnal variation of TLR. The results of this study indicated that MODIS LST can be used to calculate the near-surface TLR of grid point. The sliding window method can effectively calculate the spatially continuous TLR, and the sliding window size of 15 x 15 is suitable for this study area. The simulated TLR can be effectively corrected using the relationship between measured air temperature and LST by considering a correction coefficient. The root-mean-square error (RMSE) after correction was decreasing by 0.36 degrees C km(-1). The daily amplitude of TLR ranges from 22.25 to -13.07 degrees C km(-1), with a standard deviation (STD) of 7.50 degrees C km(-1). The near-surface TLRs vary in both space and time and are more variable than a constant of -6.5 degrees C km(-1). The mean absolute error (MAE) and RMSE between simulated hourly TLR values and measured TLR values are 2.85 and 3.32 degrees C km(-1), respectively, which means that the diurnal variation of TLR can be effectively fit using the DTC model. The research proposed method can effectively utilize MODIS LST under clear-sky conditions to calculate spatiotemporal continuous TLR.

DOI:
10.1109/TGRS.2023.3327071

ISSN:
1558-0644