Attribution of precipitation changes on ground–air temperature offset: Granger causality analysis

Abstract

This work examines the causal relationship between the value of the ground–air temperature offset and the precipitation changes for monitored 5-min data series together with their hourly and daily averages obtained at the Sporilov Geophysical Observatory (Prague). Shallow subsurface soil temperatures were monitored under four different land cover types (bare soil, sand, short-cut grass and asphalt). The ground surface temperature (GST) and surface air temperature (SAT) offset, ΔT(GST–SAT), is defined as the difference between the temperature measured at the depth of 2 cm below the surface and the air temperature measured at 5 cm above the surface. The results of the Granger causality test did not reveal any evidence of Granger causality for precipitation to ground–air temperature offsets on the daily scale of aggregation except for the asphalt pavement. On the contrary, a strong evidence of Granger causality for precipitation to the ground–air temperature offsets was found on the hourly scale of aggregation for all land cover types except for the sand surface cover. All results are sensitive to the lag choice of the autoregressive model. On the whole, obtained results contain valuable information on the delay time of ΔT(GST–SAT) caused by the rainfall events and confirmed the importance of using autoregressive models to understand the ground–air temperature relationship.