Search

Special Issue: Fog research; International audience; The aim of this article is to investigate the impact of irrigated areas on the local and regional climate in a valley in Chile's dry zone, using the KAMM model for a late spring and a late summer situation. For both situations, two different simulations are performed, featuring the actual state of the region, i.e. the valley floor used for agricultural activities with irrigation (Moist Valley Simulation MVS), and the original natural state of the region, i.e. the valley floor covered with natural vegetation (Dry Valley Simulation DVS). Agricultural activities modify the energy balance components: near-surface temperature and the thermally induced wind fields – not only in the valley, but also in some surrounding areas. With increased soil moisture and vegetation cover of the valley floor the albedo decreases and, consequently, net radiation is higher in the MVS case. An important part of the available energy is used for evaporation, and during the day the turbulent latent heat flux is higher in the MVS case; less energy is available for transformation into sensible heat flux. In the MVS case, the amplitude of the diurnal cycle of near-surface temperature on the valley floor is smaller. In the early morning, the temperature is higher while after noon it is lower in the MVS case. Outside the valley, there are zones with highly positive differences (around 2oC) which coincide with zones of positive anomalies of turbulent sensible heat flux. Near the coast, in the early morning, the intensity of the sea wind is higher while, after noon, it is lower in the MVS case. Agricultural activities enhance up-slope and upvalley winds in the daytime. With increased water availability in the soil of the valley, near-surface specific humidity is higher, too. Due to advective processes this also affects neighbouring areas. In the early morning and in the late afternoon the relative humidity in the cultivated areas reaches near-saturation values which may in fact induce fog formation at these times of the day. Nevertheless, we have to note that fog is not simulated explicitly in KAMM. Only threshold values for relative humidity have been used to indicate areas where fog formation is likely after changes in land use. In order to model fog formation more adequately, in addition to the dynamic and vegetation models, especially cloud and aerosol microphysics as well as radiation have to be considered. Efficient parameterisation schemes for radiation and cloud physics should be implemented in KAMM to represent fog conditions and their geographical distribution in more detail.