Already semi-arid, due to the effects of climate change, Botswana has been experiencing unreliable water supplies over the past several years. However, the limited climate information over different catchments makes engaging in an informed decision-making process difficult. The Notwane catchment at Gaborone dam, located in the headstreams of the Notwane River in eastern Botswana, is a major water supply for the country. However, due to the sparse network of hydrometeorological measurement stations, no reliable predictions can be made and, thus, creating a reliable runoff estimation for the reservoir has been difficult.  Through  SASSCAL, an experimental set of automated weather stations has been set up in the Notwane catchment. Preliminary analysis using artificial neural networks (ANNs) to examine the predictive capacity of the monitored variables (from July 15, 2016, through June 25, 2017: 346 days) on precipitation at four individual stations reveals that the gathered hydro-meteorological data may be useful given an increase in record length coupled with consideration of different modeling approaches to validate inherent relationships with precipitation. Study also revealed that simulated precipitation for the area exhibits similar mean and variability to the observations despite poor simulations for extreme precipitation events. These results give insight into prospects for improved hydrologic and water resource modeling over the catchment.