Agricultural management effects on mean and extreme temperature trends

Abstract

Abstract. Regression-based trend analysis is applied to observations and present-day ensemble simulations with the Community Earth System Model to assess if climate models overestimate warming trends because theoretical constant levels of irrigation and conservation agriculture (CA) are excluded. At the regional scale, an irrigation- and CA-induced acceleration of the annual mean near-surface air temperature (T2m) warming trends and the annual maximum daytime temperature (TXx) warming trends were evident. Estimation of the impact of irrigation and CA on the spatial average of the warming trends indicated that irrigation and CA have a pulse cooling effect on T2m and TXx, after which the warming trends increase at a greater rate than the control simulations. This differed at the local (subgrid) scale under irrigation where surface temperature cooling and the dampening of warming trends were both evident. As the local surface warming trends, in contrast to regional trends, do not account for atmospheric (water vapour) feedbacks, their dampening confirms the importance of atmospheric feedbacks (water vapour forcing) in explaining the enhanced regional trends. At the land surface, the positive radiative forcing signal is too weak to offset the local cooling from the irrigation-induced increase in the evaporative fraction. Our results underline that agricultural management has complex and nonnegligible impacts on the local climate and highlights the need to account for land management in climate projections.