Influence of moisture source dynamics and weather patterns on stable isotopes ratios of precipitation in Central-Eastern Africa.

We report the first δO and δH data of Virunga rainfall in the Eastern Democratic Republic of the Congo, situated on the limit between Central and Eastern Africa. The dataset is from 13 rain gauges deployed at Mount Nyiragongo and its surroundings sampled monthly between December 2013 and October 2015. The δO and δH vary from -6.44 to 6.16‰, and -32.53 to 58.89‰ respectively, and allowed us to define a LMWL of δH = 7.60δO + 16.18. Three main wind directions, i.e. NE, E and SE, were identified in the upper atmosphere corresponding to three major moisture source regions. On the contrary, lower atmospheric winds are weaker in nature and originate mainly from the S and SW, creating a topographically-driven, more local moisture regime. The latter is due to the accumulation in the floor of the rift of water vapor from Lake Kivu forming a layer of isotopically enriched vapor that mediates the isotope enrichment of the falling raindrops. A strong seasonality is observed in both δO and δH data, and is primarily driven by combined seasonal and spatial variation in the moisture sources. The δO and δH seasonality is thus correlated to weather patterns, as the latter control the wet to dry season shifting, and vice versa. The key characteristic of seasonality is the variation of monthly precipitation amounts, since the mean monthly air temperature is nearly constant on an annual scale. Two regionally relevant hydrological processes contribute to the isotopic signature: namely moisture uptake from the isotopically enriched surface waters of East African lakes and from the depleted soil-water and plants. Consequently, the proportion of water vapor from each of these reservoirs in the atmosphere drives the enrichment or depletion of δH and δO in the precipitation. Thus, during wet periods the vapor from soil-plants evapotranspiration dominates yielding isotopically depleted precipitation, contrary to dry periods when vapor from lakes surface evaporation dominates, yielding isotopically enriched precipitation. At the global scale, our dataset reduces gaps in this region that has been poorly studied for δO and δH in precipitation. At the regional scale, the improved understanding of the ways land cover, moisture source seasonal and spatial dynamics, and atmospheric patterns impact precipitation spatial and temporal variabilities in Central-East African will contribute to the ongoing research on mitigating the impacts of ongoing climate change in Sub-Saharan Africa. The reduction of gaps and uncertainties in δH and δO of precipitation, and the understanding of their interrelation with weather patterns are essential for a better past, present and future environmental and climatic modelling at both local and regional scales.