Fine-scale mapping of daily minimum temperature in a cropland with complex terrains through the combination of a cold flow accumulation model with inversion strength

Abstract

In complex terrains, the minimum temperature varies at fine spatial scales owing to cold air drainage. This leads to large spatial variabilities in phenological events and frost risk in plants. Various topographic variables have been used to model cold air drainage and its influence on temperature. In this study, we modified a conventional approach using flow accumulation (FA). First, we developed a new algorithm for FA by reconsidering effective areas of cold air drainage. This allowed us to increase the model's applicability. The FA model was then combined with the vertical gradient of potential temperature, which represents the inversion strength during stable, nighttime conditions. This process facilitated minimum temperature estimation at a given date and point. The model was developed using temperature observations in tea fields with complex terrains, where frost risk is increasing due to climate change. The maximum difference in daily minimum temperature across 15 observation points was 6.7°C, despite only a 73-m elevation difference. Compared with models using spatial interpolation alone, the proposed model produced a better estimate of the spatiotemporal variability of daily minimum temperature and improved the average residuals between estimated and measured temperatures in valley areas from 1.9 to 0.03°C. Moreover, compared to the conventional algorithm, the proposed algorithm improved temperature estimations in ridge areas and clarified the boundaries between ridges and valleys. Our approach could contribute to the spatiotemporal analysis of temperature variations and concomitant changes in phenological events and frost risk in complex terrains.