The industrialization process taking place in Africa has led to an overall increase in groundwater abstraction in most countries in the continent. However, the lack of hydrogeological data, as in many developing countries, makes it difficult to properly manage groundwater systems. This study presents a real case study in which a combination of different hydrogeological tools together with different sources of information allow the assessment of how increased competition for water may be affecting groundwater systems by analysing the sustainability of new abstraction regimes under different real climatic condition (before, during and after La Niña 2016). The area where this approach has been applied is Kwale County (in Coastal Kenya) in a hydrogeological context representative of an important part of the east coast of the continent, where new mining and agriculture activities co-exist with tourism and local communities. The results show that the lack of aquifer systems data can be overcome, at least partly, by integrating different sources of information. Most of the time, water-reliant users collect specific hydrogeological information that can contribute to defining the overall hydrogeological system, since their own main purpose is to exploit the aquifer with the maximum productivity. Therefore, local community water usage, together with different stakeholder's knowledge and good corporate water management act as a catalyst for providing critical data, and allows the generation of credible models for future groundwater management and resource allocation. Furthermore, complementary but simple information sources such as in situ interviews, Google Earth, Trip Advisor and easy-to use analytical methods that can be applied in the African context as in many developing countries, and enables groundwater abstraction to be estimated and the sustainability of the aquifer system to be defined, allowing potential future risks to be assessed.

In the context of developing countries, effective groundwater resource management is often hindered by a lack of data integration between resource availability, water demand, and the welfare of water users. As a consequence, drinking water-related policies and investments, while broadly beneficial, are unlikely to be able to target the most in need. To find the households in need, we need to estimate their welfare status first. However, the current practices for estimating welfare need a detailed questionnaire in the form of a survey which is time-consuming and resource-intensive. In this work, we propose an alternate solution to this problem by performing a small set of cost-effective household surveys, which can be collected over a short amount of time. We try to compensate for the loss of information by using other modalities of data. By combining different modalities of data, this work aims to characterize the welfare status of people with respect to their local drinking water resource. This work employs deep learning-based methods to model welfare using multi-modal data from household surveys, community handpump abstraction, and groundwater levels. We employ a multi-input multi-output deep learning framework, where different types of deep learning models are used for different modalities of data. Experimental results in this work have demonstrated that the multi-modal data in the form of a small set of survey questions, handpump abstraction data, and groundwater level can be used to estimate the welfare status of households. In addition, the results show that different modalities of data have complementary information, which, when combined, improves the overall performance of our ability to predict welfare.