Yes, I do agree.

Nowadays, there are lots of readings from sensors spread out all over the Planet alongside with relevant info (orography, hydrography) and historic records. Cloud services provide the ability to collect, store and process this ginormous amount of data. Proper computation approaches, such as tensor based computation, permit to effectively compute and model these intractably big data. Additionally, and in my opinion, atmospheric events, climate changes or water availability follow patterns that can be accurately modeled by state of the art machine learning (AI) algorithms, making prediction possible, into some extent.

In today's time, there is a huge volume of data being collected on climate, hydrology and other related aspects. To make sense of this data requires computing power that only Computers can provide. Therefore model building and scenario analysis are ways of doing this. They may or may not be Artificial Intelligence, which is a loosely used terminology.

With large datasets from monitoring of environmental parameters, we are facing big data that spans over time and geographic space. AI is a perfect fit to process these datasets and look for patterns in the data that can guide decisions in environmental sciences. Some examples of data that is amenable to AI are: streamflow time series, precipitation patterns, soil moisture fluxes, contamination transfer, temperature variability, remote-sensing of patterns, visualization, and environmental impact studies. Neural network models and other AI methods in identifying functional patterns from GIS datasets and large temporal datasets is emerging as important in environmental and hydrologic sciences.