This paper reviews the current GeoAI and machine learning applications in hydrological and hydraulic modeling, hydrological optimization problems, water quality modeling, and fluvial geomorphic and morphodynamic mapping. GeoAI effectively harnesses the vast amount of spatial and non-spatial data collected with the new automatic technologies. The fast development of GeoAI provides multiple methods and techniques, although it also makes comparisons between different methods challenging. Overall, selecting a particular GeoAI method depends on the application’s objective, data availability, and user expertise. GeoAI has shown advantages in non-linear modeling, computational efficiency, integration of multiple data sources, high accurate prediction capability, and the unraveling of new hydrological patterns and processes. A major drawback in most GeoAI models is the adequate model setting and low physical interpretability, explainability, and model generalization. The most recent research on hydrological GeoAI has focused on integrating the physical-based models’ principles with the GeoAI methods and on the progress towards autonomous prediction and forecasting systems.

The patchy distribution of high-Andean treeline forests has provoked discussion about the relative importance of anthropogenic and climatic causes of this pattern, both of which vary with topography. We aimed to understand the topographic controls on the distribution of Polylepis subsericans treeline forests in the Andes of southern Peru, and the changes in these controls along an elevational gradient. We mapped Polylepis forests in the Cordillera Urubamba, Cusco, using high-resolution aerial images and related forest cover to topographic variables extracted from a digital terrain model (30-m resolution). The variables were selected based on their expected biological relevance for tree growth at high elevations. We constructed logistic regression models of forest cover, separately for each of five 100-m elevational belts. To deal with spatial autocorrelation, models were based on randomized 10% subsampling of the data with 1000 repetitions. The results suggest a consistent shift in topographic preference with elevation, with forests at lower elevations showing a preference for topographically protected sites near rivers and forests at higher elevations being increasingly restricted to north-facing and well-drained sites. Our study offers the first indication of the ability of Andean treeline forests to benefit from the topographic heterogeneity of the high-Andes. Providing that dispersal and establishment are possible, local relocation between microsites could help these forests to persist regionally in spite of changing climatic conditions.

Increased exposure to greener environments has been suggested to lead to health benefits in children, but the associated mechanisms in early life, particularly via biological mediators such as altered maternal milk composition, remain largely unexplored. We investigated the associations between properties of the mother’s residential green environment, measured as (1) greenness (Normalized Difference Vegetation index, NDVI), (2) Vegetation Cover Diversity (VCDI) and (3) Naturalness Index (NI), and human milk oligosaccharides (HMOs), known for their immune- and microbiota-related health effects on the infant (N = 795 mothers). We show that HMO diversity increases and concentrations of several individual HMOs and HMO groups change with increased VCDI and NI in residential green environments. This suggests that variation in residential green environments may influence the infant via maternal milk through modified HMO composition. The results emphasize the mediating role of breastfeeding between the residential green environments and health in early life.