The Congo Basin exhibits tremendous heterogeneities, out of which it emerges as an intricate system where complexity will vary consistently over time and space. Increased complexity in the absence of adequate knowledge will always result in increased uncertainties. One way of simplifying this complexity is through an understanding of organisational relationships of the landscape features, which is termed here as catchment classification. The need for a catchment classification framework for the Congo Basin is obvious given the basin’s inherent heterogeneities, the ungauged nature of the basin, and the pressing needs for water resources management that include the quantification of current and future supplies and demands, which also encompass the impacts of future changes associated with climate and land use, as well as water resources operational policies. The need is also prompted by many local-scale management concerns within the basin. This study uses an a priori approach to determine homogenous climatic-physiographic regions that are expected to underline dominant hydrological processes characteristics. A set of 1740 catchment units are partitioned across the whole basin, based on a set of comprehensive criteria, including natural break of the elevation gradient (199 units), inclusion of socio-economic and anthropogenic systems (204 units), and water management units based on traditional nomenclature of the rivers within the basin (1337 units). The identified catchment units are used to assess existing datasets of the basin physical properties, necessary to derive descriptors of the catchments characteristics. An unsupervised classification, based on Hierarchical Agglomerative Cluster algorithm is used, that yields 11 homogenous groups that are consistent with the current perceptual understanding of the Congo Basin physiographic and climatic settings. These regions represent therefore an a priori classification that will be further used to derive functional relationships of the catchments, necessary to enable hydrological prediction and water management in the basin.

Hydroclimatic flood generating processes, such as excess rain, short rain, long rain, snowmelt and rain-on-snow, underpin our understanding of flood behaviour. Knowledge about flood generating processes helps to improve modelling decisions, flood frequency analysis, estimation of climate change impact on floods, etc. Yet, not much is known about how climate and catchment attributes influence the distribution of flood generating processes. With this study we aim to offer a comprehensive and structured approach to close this knowledge gap. We employ a large sample approach (671 catchment in the conterminous United States) and test attribute influence on flood processes with two complementary approaches: firstly, a data-based approach which compares attribute probability distributions of different flood processes, and secondly, a random forest model in combination with an interpretable machine learning approach (accumulated local effects). This machine learning technique is new to hydrology, and it overcomes a significant obstacle in many statistical methods, the confounding effect of correlated catchment attributes. As expected, we find climate attributes (fraction of snow, aridity, precipitation seasonality and mean precipitation) to be most influential on flood process distribution. However, attribute influence varies both with process and climate type. We also find that flood processes can be predicted for ungauged catchments with relatively high accuracy (R2 between 0.45 and 0.9). The implication of these findings is that flood processes should be taken into account for future climate change impact studies, as impact will vary between processes.