This paper provides a simple methodology to detect long-term trends in groundwater resources by using multi-date remotely sensed imagery and in-situ observations to monitor transitions in vegetation in and around Lake Ngami's emergent floodplain environment in the distal reaches of Botswana's Okavango Delta. Covering a period of 34 years between 1967 and 2001, trends emerging from this reconstruction point to a sustained shift from a perennial wetland to an intermittently flooded dryland environment, significant increase in drought tolerant woody species notably Acacia mellifera and Acacia erioloba, and sustained contraction of groundwater resources. With transitions in the distribution of these and other indicator species mimicking changes in hydrological conditions, it is apparent that spatial and temporal variations in their distribution can be used to provide long-term trends in groundwater resources. In view of the general lack of time-series data on regional trends in aquifer storages in arid and semi-arid areas, remote-sensing-based monitoring of vegetation can be exploited to provide proxy measurements and useful insights that can be used to guide the formulation of informed interventions potentially capable of enhancing our capacities to cope with decreasing supplies in areas where signals of climate change point to the inevitability of increased scarcities. Key words: groundwater, long-term trends, remote sensing, arid/semi arid areas, drying sequences.