Understanding soil moisture - precipitation coupling on mesoscales using observations over North Africa

Soil moisture is a main driver of land-atmosphere interactions. On the meso-scale (10-100 km), soil moisture in uences the formation and development of rainfall systems, with at- tendant consequences for the hydrological, agricultural and social-economic sectors. Under- standing of the strength and the sign of the soil moisture { precipitation coupling (SMPC) has a number of relevant implications for the present and future climates. To date, there is signicant uncertainty in how soil moisture aects rainfall, owing largely to the lack of observational evidence. In this thesis, analysis of 10 years of satellite soil moisture and pre- cipitation records using a common analysis framework is done to advance our understanding of the SMPC from an observational perspective in the North African region. The present study reproduces, evaluates and applies the statistical method of Taylor et al. (2012), which links the probability of afternoon rainfall occurrence to the preceding soil moisture conditions. The results of this evaluation indicate that the estimated coupling relationship largely relies on the set up of introduced assumptions and limitations. A number of assumptions used by the method maximize the probability of the negative spatial SMPC in the region. The sensitivity tests identify the following major conditions which maximize the likelihood of negative coupling: (i) - the afternoon accumulated precipitation (AAP) threshold value used in the method to dene a rainfall event should be between 1 and 15 mm/ 9hrs, (ii) - the areal distribution of the AAP should be taken into account, and (iii) - events with smaller AAP area should be selected. The latter condition is not considered in the current set up. Except for that, the original framework is found to be optimal for its purposes. The preference of rainfall to occur over spatially drier soils in the region is found to be robust. The above ndings together with the simplicity of the framework can be identied as a key for its applicability and further improvement. The method is then applied to investigate the strength and spatial variability of the SMPC, as well as the temporal eects of soil moisture on moist convection. The analyses show that the negative spatial and the negative temporal coupling relationships co-exist and are not independent of one another. Regional application of the method at a higher than before horizontal resolution of 1 reveals previously averaged eects of wetland and irrigated land areas on the SMPC measure, and identies regions, where surface eects on rainfall can be expected to be particularly strong. These are the south-western part of the domain with the most robust negative SMPC signal, and the South Sudan. In the latter region, the spatial coupling is found to be modulated by the presence of wetlands and is susceptible to the amount of long-lived propagating convective systems. In the drier northern latitudes the low-level atmospheric moisture variability is suggested to be more decisive than the soil moisture state. The results of this thesis provide the rst insight into the regional variability and potential "hot-spots" of the SMPC in North Africa with further implications for the hydrological, agricultural, economical sectors, as well as climate change analyses and planning of future measurement campaigns. Though focused on North Africa, the results of this study are relevant and are likely to be applicable to other semi-arid climates