Using stable isotopes to inform water resource management in forested and agricultural ecosystems.

Present and future climatic trends are expected to markedly alter water fluxes and stores in the hydrologic cycle. In addition, water demand continues to grow due to increased human use and a growing population. Sustainably managing water resources requires a thorough understanding of water storage and flow in natural, agricultural, and urban ecosystems. Measurements of stable isotopes of water (hydrogen and oxygen) in the water cycle (atmosphere, soils, plants, surface water, and groundwater) can provide information on the transport pathways, sourcing, dynamics, ages, and storage pools of water that is difficult to obtain with other techniques. However, the potential of these techniques for practical questions has not been fully exploited yet. Here, we outline the benefits and limitations of potential applications of stable isotope methods useful to water managers, farmers, and other stakeholders. We also describe several case studies demonstrating how stable isotopes of water can support water management decision-making. Finally, we propose a workflow that guides users through a sequence of decisions required to apply stable isotope methods to examples of water management issues. We call for ongoing dialogue and a stronger connection between water management stakeholders and water stable isotope practitioners to identify the most pressing issues and develop best-practice guidelines to apply these techniques.

Machine Learning as a Strategic Tool for Helping Cocoa Farmers in Côte D'Ivoire.

Machine learning can be used for social good. The employment of artificial intelligence in smart agriculture has many benefits for the environment: it helps small farmers (at a local scale) and policymakers and cooperatives (at regional scale) to take valid and coordinated countermeasures to combat climate change. This article discusses how artificial intelligence in agriculture can help to reduce costs, especially in developing countries such as Côte d'Ivoire, employing only low-cost or open-source tools, from hardware to software and open data. We developed machine learning models for two tasks: the first is improving agricultural farming cultivation, and the second is water management. For the first task, we used deep neural networks (YOLOv5m) to detect healthy plants and pods of cocoa and damaged ones only using mobile phone images. The results confirm it is possible to distinguish well the healthy from damaged ones. For actions at a larger scale, the second task proposes the analysis of remote sensors, coming from the GRACE NASA Mission and ERA5, produced by the Copernicus climate change service. A new deep neural network architecture (CIWA-net) is proposed with a U-Net-like architecture, aiming to forecast the total water storage anomalies. The model quality is compared to a vanilla convolutional neural network.

A stochastic cellular automaton model to describe the evolution of the snow-covered area across a high-elevation mountain catchment.

Variations in the extent and duration of snow cover impinge on surface albedo and snowmelt rate, influencing the energy and water budgets. Monitoring snow coverage is therefore crucial for both optimising the supply of snowpack-derived water and understanding how climate change could impact on this source, vital for sustaining human activities and the natural environment during the dry season. Mountainous sites can be characterised by complex morphologies, cloud cover and forests that can introduce errors into the estimates of snow cover obtained from remote sensing. Consequently, there is a need to develop simulation models capable of predicting how snow coverage evolves across a season. Cellular Automata models have previously been used to simulate snowmelt dynamics, but at a coarser scale that limits insight into the precise factors driving snowmelt at different stages. To address this information gap, we formulate a novel, fine-scale stochastic Cellular Automaton model that describes snow coverage across a high-elevation catchment. Exploiting its refinement, the model is used to explore the interplay between three factors proposed to play a critical role: terrain elevation, sun incidence angle, and the extent of nearby snow. We calibrate the model via a randomised parameter search, fitting simulation data against snow cover masks estimated from Sentinel-2 satellite images. Our analysis shows that.