Ecological Forecasting and Ecosystems Modeling - A Challenge to the High Performance Computing Community

As stated in the AMS Statement on The Future of Ecological Forecasting. “Environmental change is the defining challenge of our time. The unprecedented nature of this change means that we can no longer rely on historical patterns to inform environmental decision making. Ecological forecasts respond to this challenge by predicting likely changes in ecosystems and the “ecosystem services” they provide, such as food, water, fuel, security, shelter, and medicine, in response to natural and human-induced drivers.” Ecological forecasting based on ecosystem models is an important capability that we – as individuals and as society – need in order to make informed decisions as we manage our natural resources and indeed in our day to day lives.

Advances in data availability, scientific understanding, computing power, and model sophistication have enabled advances in ecological forecasting, but much more needs to be done and many challenges have to be met to continue to progress. This presentation will summarize the nature of one of those challenges – providing the high performance computing (HPC) and related capabilities needed to develop and deliver next generation ecological forecasts. The HPC challenges include the sheer computing power needed, delivering of this power to researchers and for operational ecological forecasting, and developing the software infrastructure to integrate the many existing ecosystem models and ecological forecasting codes into a cohesive framework that leverages the work that has been done to move into the future. We will provide examples of today’s state of the art, specific gaps in capability, and identify agencies and programs working in this area today. And we challenge the HPC Community to help us address the vital societal need for ecological forecasts.