Computationalizing landscapes: The CoRE stack data structure for social-ecological diagnostics and planning

The CoRE Stack (Commoning for Resilience and Equality) is a community-based digital public infrastructure consisting of datasets, algorithm pipelines, and user-facing tools, that can be used by rural communities and other stakeholders to improve the sustainability and resilience of their local landscapes. It brings together over a hundred variables (and growing) to describe landscape units like micro-watersheds in terms of changes in the area over the years in cropping intensity, water-table levels, health of waterbodies, forests and plantations, and welfare fund allocation, among others, sourced from multiple contributors or built using open ML models operating on satellite data. The unique aspect with the CoRE stack is that all this is assembled in a neat entity-relation data structure which describes landscapes in terms of discrete spatial units with nesting, intersection, and other relationships between them. This makes it computationally accessible to slice and dice the data like we would in a data frame, to understand landscapes more closely. We will present this data structure and the remarkable agility it unlocks in building new applications on top of it. We will demonstrate this through a novel application, CoRE insights, which uses the CoRE stack data on the one side, and extracts landscape diagnostic rulesets using LLMs from research papers on the other side, to build a neurosymbolic framework that can apply these rulesets to diagnose specific problems in a landscape, like whether or not it is suffering from groundwater stress, or rainfed agricultural risk, and so on, with uncertainty information on this determination. Another layer, the CoRE lens, is a library we are developing to interact with the data more easily using a functional programming approach which keeps the code short and easy to debug and understand. We will conclude with real stories of change on the ground using the CoRE stack, where the tools are in use by landscape stewards in 1500+ villages in India to build water security and ecological restoration plans with communities.

We will try to drive home the point that in this increasingly data-crazy world, it is important to get underlying data structure right, to be able to leverage the opportunities in front of us now.