Development of High Resolution and Homogenized Gridded Land Surface Air Temperature Data: A Case Study Over Pan-East Asia

Jiayi Cheng,Qingxiang Li,Phil Jones,Liya Chao,Suman Maity,Boyin Huang

doi:10.3389/fenvs.2020.588570

Abstract

The Land Surface Air Temperature (LSAT) climatology during the period of 1961-1990 and the anomalies (relative to the 1961-1990 climatology) have been developed over Pan-East Asian region at a (monthly) 0.5°×0.5° resolution. The development of these LSAT data sets are based on the recently released C-LSAT station datasets and the high resolution Digital Elevation Model (DEM), and interpolated by the Thin Plate Spline (TPS) method (through ANUSPLIN software) and the Adjusted Inverse Distance Weighting (AIDW) method. Then they are combined into the high resolution gridded LSAT datasets (including the monthly mean, maximum, and minimum temperature). Considering the mean LSAT for example, the Cross Validation (CV) of the datasets indicates that the regional average of the Root Mean Square Error (RMSE) for the climatology is about 0.62℃, and the average RMSE and Mean Absolute Error (MAE) for the anomalies are between 0.47 - 0.90 ℃ and 0.32 - 0.63 ℃ during the study period. The analysis also demonstrate that the gridded anomalies describe the spatial pattern fairly well for the coldest (1912, 1969) and the warmest (1948, 2007) years during the first and second half of the 20th century. Further analysis reveals that the high resolution dataset also performs well in the estimation of long-term LSAT change trend. Thus it can be concluded that this newly constructed datasets is a useful tool for regional climate monitoring, climate change research as well as climate model verification.

Highlights

Land Surface Air Temperature (LSAT) is considered one of the important indicators of the global and regional climate change
We first split the land surface air temperature from C-LSAT2.0 into climatology and anomaly data, and interpolate them into a 0.5◦ × 0.5◦ grid datasets, respectively. They are combined into a new monthly high-resolution LSAT grid data set for the period of 1900–2018 over Pan-East Asia
The validation of the dataset shows that the interpolated dataset is of good accuracy in both climatology and the LSAT anomaly variations

Summary

Introduction

Land Surface Air Temperature (LSAT) is considered one of the important indicators of the global and regional climate change. Regional High Resolution Gridded LAST or continental regions at century scale To resolve these issues, scientists developed homogenized station datasets and converted them into gridded datasets for the convenience of applications (Hutchinson, 1991; Daly et al, 1994; Li and Li, 2007; Xu et al, 2009). For global large-scale climate temperature trend estimation using low-resolution datasets (normally in 5◦ × 5◦ resolution to avoid changes at small scales) (Jones and Briffa, 1992; Peterson and Vose, 1997; Hansen et al, 1999; Li et al, 2017; Xu C. et al, 2018; Yun et al, 2019) can basically meet the accuracy requirements. A consistent outcome from available low-resolution data sets is that the global land temperature trend since 1880 has become more and more significant (Li et al, 2020)

Objectives

Methods

Conclusion