Machine learning (ML) applications in Earth and environmental sciences (EES) have gained incredible momentum in recent years. However, these ML applications have largely evolved in ‘isolation’ from the mechanistic, process-based modelling (PBM) paradigms, which have historically been the cornerstone of scientific discovery and policy support. In this perspective, we assert that the cultural barriers between the ML and PBM communities limit the potential of ML, and even its ‘hybridization’ with PBM, for EES applications. Fundamental, but often ignored, differences between ML and PBM are discussed as well as their strengths and weaknesses in light of three overarching modelling objectives in EES, (1) nowcasting and prediction, (2) scenario analysis, and (3) diagnostic learning. The paper ponders over a ‘coevolutionary’ approach to model building, shifting away from a borrowing to a co-creation culture, to develop a generation of models that leverage the unique strengths of ML such as scalability to big data and high-dimensional mapping, while remaining faithful to process-based knowledge base and principles of model explainability and interpretability, and therefore, falsifiability.

Understanding of hydroclimatic processes in Africa has been hindered by the lack of in situ precipitation measurements. Satellite-based observations, in particular, the TRMM Multisatellite Precipitation Analysis (TMPA) have been pivotal to filling this void. The recently released Integrated Multisatellite Retrievals for GPM (IMERG) project aims to continue the legacy of its predecessor, TMPA, and provide higher-resolution data. Here, IMERG-V04A precipitation data are validated using in situ observations from the Trans-African Hydro-Meteorological Observatory (TAHMO) project. Various evaluation measures are examined over a select number of stations in West and East Africa. In addition, continent-wide comparisons are made between IMERG and TMPA. The results show that the performance of the satellite-based products varies by season, region, and the evaluation statistics. The precipitation diurnal cycle is relatively better captured by IMERG than TMPA. Both products exhibit a better agreement with gauge data in East Africa and humid West Africa than in the southern Sahel. However, a clear advantage for IMERG is not apparent in detecting the annual cycle. Although all gridded products used here reasonably capture the annual cycle, some differences are evident during the short rains in East Africa. Direct comparison between IMERG and TMPA over the entire continent reveals that the similarity between the two products is also regionally heterogeneous. Except for Zimbabwe and Madagascar, where both satellite-based observations present a good agreement, the two products generally have their largest differences over mountainous regions. IMERG seems to have achieved a reduction in the positive bias evident in TMPA over Lake Victoria.