Use Case 7

This use case focuses on the development of automated tools for impact assessment of climate events.

Partner:

Leibniz Centre for Agricultural Landscape Research (ZALF)
TUM School of Life Sciences (TUM)
Deutscher Wetterdienst (DWD)
German Climate Computing Centre (DKRZ)

Description

Motivation:

Climate change is increasingly recognized as one of the greatest threats to global agriculture, with rising temperatures, altered precipitation patterns, and the frequency of extreme weather events all jeopardizing crop production. To prepare for and mitigate these impacts, it is essential to conduct comprehensive climate impact assessments that can accurately project agricultural yield at multiple spatial and temporal scales. However, a critical barrier to these assessments is the fragmentation of climate and weather data, which are essential for understanding the scope of climate change impacts.

Currently, climate data is scattered across different infrastructures, with significant variations in spatial resolution, temporal coverage, and intended use. For example, weather datasets are often optimized for short-term forecasting, while climate models such as CMIP6 and CMIP5 focus on long-term projections, but these models are not directly comparable due to differences in their temporal and spatial scales. This fragmentation severely limits the ability to synthesize data for comprehensive agricultural studies. Given the growing urgency of addressing climate-related risks to agriculture, there is an immediate need to bridge this gap and provide researchers with the tools necessary to harmonize climate data and assess its implications for crop production.

The challenge lies not just in analyzing climate scenarios but in effectively integrating fragmented datasets across different domains, time periods, and spatial resolutions. This is critical for developing actionable insights that can guide agricultural adaptation strategies and ensure food security in a rapidly changing climate.

Objectives:

This use case aims to overcome the challenge of fragmented climate data by developing advanced digital tools and algorithms that enable a unified workflow for processing and using bias-corrected, downscaled climate change scenarios. These tools will facilitate agricultural impact assessments by offering consistent, harmonized datasets at various spatial resolutions.

Objectives include:

1. Data Integration: Develop algorithms that harmonize and integrate data from different climate models (e.g., CMIP6 vs. CMIP5) to create a cohesive set of climate scenarios. These datasets will be used to assess impacts on agricultural systems, particularly wheat yields, across different spatial scales (from plot-level to regional scales).
2. Impact Comparison: Use hybrid models to assess how wheat yields are impacted by different climate scenarios and how these impacts have evolved over time. This comparison will examine past, present, and projected climate scenarios to gauge the future trajectory of agricultural vulnerabilities and opportunities.
3. Automated Data Workflow: Create a user-friendly, automated dashboard for researchers, enabling seamless data extraction, model selection, visualization, and analysis. This tool will empower agricultural researchers to easily access and analyze climate scenario data, thereby streamlining climate and crop impact assessments.
4. Integration of Large Language Models (LLMs): Leverage LLMs for querying databases using natural language, facilitating data accessibility and making complex climate data more understandable and usable for agricultural scientists and researchers across domains (e.g., AgMIP, ISIMIP, AgML).

Actions:

1. Development of Spatio-temporally Harmonized Downscaled High Resolution Climate datasets
2. Automated Dashboard for Data Access and Analysis
3. Proof-of-Concept for LLM Integration for querying and download

More Use Cases