Modeling and analyzing the effects and feedbacks of irrigation on land-atmosphere interaction - Using a regional climate model up to convection-permitting scale

Abstract

With land use practices humans alter the biophysical and biogeochemical properties of the land surface. A widely applied land use practices in agriculture is irrigation, which aims for improved growth conditions of crops by increasing the soil moisture. Multiple studies have investigated irrigation effects on both the global and regional scale. Often these studies focus on large-scale irrigated areas such as in India, China or in the US. However, there is a lack of studies investigating the effects of small-scale irrigation, particularly in regions with heterogeneous land cover such as Europe. Here, regional climate models (RCMs) have the advantage of their high spatial resolution. Within the last decade, various RCMs have received irrigation parameterizations, which are governed by the model’s characteristics. Against this background, this cumulative dissertation targets at the development, implementation and application of a novel irrigation parameterization for the new version of the RCM REMO2020, interactively coupled to its vegetation module iMOVE. This model setup enables the interaction of irrigation effects and feedbacks between soil, atmosphere and vegetation processes. With the advancements in climate modeling towards higher resolutions, the parameterization should be applicable to high resolution. Therefore, a subgrid-scale approach with a separate irrigated fraction is realized, which represents the heterogeneous soil moisture distribution caused by irrigation.The newly developed parameterization increases the soil moisture directly if irrigation is required during the growing season. Irrigation requirement is assessed based on a user-defined irrigation threshold. For the water application three different schemes can be selected, depending on the research aim. The irrigation parameterization is applied and evaluated in two consecutive, largely complementary simulation studies. In the first study, the newly developed irrigation parameterization is applied at 0.11° horizontal resolution for the case study area of South-Western Europe, with a focus on the Po Valley in Northern Italy as one of the most irrigated areas in Europe. Reanalysis-driven simulations are conducted for the year 2017, a year characterized by multiple heatwaves, with and without the irrigation parameterization. The application of the irrigation parameterization with the consequent increase of soil moisture causes effects and feedbacks on land, atmosphere and vegetation. For example, the surface energy balance shows an increased latent heat flux and a decreased sensible heat flux. Furthermore, higher evapotranspiration rates increase the 2 m relative humidity and lowers the 2 m mean temperature. The results indicate that irrigation has the ability to reduce the intensity of heatwaves. Vegetation processes strongly depend on the soil moisture and the 2 m temperature. The results show the LAI responds to irrigation with a slower growth, but a higher LAI peak. The second study employs the non-hydrostatic version of REMO2020-iMOVE with the irrigation parameterization at convection-permitting scale (0.0275°) for a case study around the Po Valley. The simulations are nested into the simulations at 0.11° horizontal resolution from the first study, using the same irrigation settings and the same period, making them comparable. The higher resolution at 0.0275° allows for a more accurate representation of land surface features, such as topography and irrigated areas, resulting in more grid cells with higher irrigated fractions. This leads to more localized and more pronounced air temperature effects at 0.0275° horizontal resolution. Irrigation effects on vegetation develop similarly at both resolutions. However, the interactive coupling makes them sensitive to changes in atmospheric conditions due to the different resolutions. In particular, precipitation is represented differently at both resolutions. The diurnal cycle of precipitation is improved at convection-permitting resolution compared to observational values. Furthermore, irrigation effects on precipitation develop very different at 0.11° and at 0.0275° horizontal resolution. While at 0.11° horizontal resolution, irrigation leads to precipitation increase at the border of the Alps, but inhibits convection above irrigated areas, at 0.0275° the effects on precipitation are characterized with small-scale features and mixed signals. In total, at convection-permitting scale irrigation leads to a precipitation reduction.