Reliable recharge assessment for sound hydrogeological analysis in arid and semi-arid area is a challenge due to the complex process requiring of intensive data mining and especially in face of limited access to data sets or only short data periods available. This has forced researchers to improve the available data using remote sensing products for input data (e.g., TRMM, CMORPH, TMPA). Recharge estimation through a water-soil-plant distributed model based on satellite products appears to be an interesting solution being adopted at the Lake Chad Basin for the Maiduguri area (Southwest of Lake Chad basin). Natural recharge is performed with a water-soil-plant distributed model (Visual Balan) based in meteorological data obtained from a ground station (Miduguri, 532 mm) and from satellite products (MSWEP, 673 mm) for the same geographic location of ground station and satellite for the 2009-2010 period. Satellite products generally overestimate precipitation with regard to ground station; this fact is attributed to no raining cirrus with cold cloud-top temperatures. As a result, groundwater recharge estimation in the area by field stations and satellite products is 80 and 100 mm per year respectively, a 20% of difference in the final value. Considering the non-linearity of aquifer recharge process and dependence on precipitation values, comparison between ground-based and satellite data sets is the initial step for recharge estimation assessment.