Building a Drought Planning Platform

Drought is one of the most costly weather and climate disasters in the United States and poses a significant threat to the economic and social wellbeing of communities across the Nation. Planning for future droughts can support better informed decisions, leading to reduced impacts and costs. 

NOAA’s National Integrated Drought Information System (NIDIS), in partnership with the U.S. Bureau of Reclamation, is building a Drought Planning Platform on Drought.gov. 

The Drought Planning Platform will coordinate the best available trusted public data and resources on drought planning, leveraging artificial intelligence to provide science-based technical assistance and tailored guidance that supports Western U.S. communities in reducing future drought risk and impacts.

The Platform empowers users to develop, update, and communicate their own locally-relevant drought plan by guiding them through the process, recommending evidence-based approaches, and building their capacity to justify and maintain their decisions over time.

The Platform design and content is based on documented needs of drought and water planners. A prototype is under development with an initial focus on the Four Corners states (Arizona, Colorado, New Mexico, Utah). A thorough user feedback process will be utilized to improve the prototype before expanding to the broader Western U.S.  

The Platform will address five key steps for developing a comprehensive drought plan:

• Where do I start in developing a plan? This includes guidance for developing an organizational structure and identifying key roles and functions.
• How will we recognize drought in the early stages? This includes understanding how to measure and monitor drought, selecting key indicators and indices, selecting drought stages and triggers, and communicating drought conditions.
• How will drought affect us? This includes understanding historical drought context and impacts, and conducting your own assessment to understand your vulnerability to current and future droughts.
• How can we protect ourselves from the next drought? This includes identifying and prioritizing both response actions during a drought and mitigation actions before future droughts, and communicating drought actions.
• How do I implement and keep the plan relevant? This includes guidance to update and evaluate the plan.

Harnessing generative AI to advance service delivery for drought planning:

This  first-of-its-kind decision support tool will leverage artificial intelligence to rapidly access, interpret, and visualize authoritative information tailored for specific drought plans. The Platform will be hosted on Drought.gov and powered by the Argonne Resilience AI Assistant (ARAIA) from Argonne National Laboratory, a large language model (LLM)-integrated decision support framework developed specifically to help stakeholders understand natural hazard risk and associated operational challenges, as well as identify relevant adaptation strategies to address these challenges. ARAIA incorporates advanced AI methodologies, including retrieval augmented generation (RAG)-based multi-agent LLM systems, to integrate data from scientific literature, simulations, and observational sources. ARAIA additionally implements strict guardrails to ensure operational security and reliability while preventing bias. 

The AI component of the Drought Planning Platform is being designed as an intuitive decision support system, allowing users to ask questions and obtain clear, fact-based responses that are useful for developing drought plans. Users will be able to interact with the system to retrieve specific data, such as drought indices and drought vulnerability information most relevant for their location, and receive actionable recommendations on areas such as drought mitigation planning and potential actions to take during a drought to minimize risk. All content will be evaluated by subject matter experts, ensuring its reliability for resilience strategies.

For additional information, please contact Meredith Muth (meredith.f.muth@noaa.gov)