ClimData: Enabling User-Friendly Access to Climate Data for Agriculture (UC7)

ClimData: Enabling User-Friendly Access to Climate Data for Agriculture (UC7) Kaushik Muduchuru, Amit Srivastava, Frank Ewert Multi-Scale Modelling & Forecasting, ZALF Leibniz Centre for Agricultural Landscape Research (ZALF) | Eberswalder Straße 84 | 15374 Müncheberg | Germany Contact: Kaushik Muduchuru ([email protected]) Date: 30.09.2025 1. Automated climate & weather data extraction Automated pipeline to fetch, extract climate, quality check and downscale weather data. Data sources: CMIP6, NASA NEX-GDDP, MSWX, DWD HYRAS, ERA5/ERA5-Land (multi-source reanalysis & model projections) Configurable workflowsData Access 2. Core Functionalities Data Access – Unified retrieval from multiple climate datasets with provenance tracking. Reproducibility – Hydra-based configurations ensure repeatable and scalable workflows. Standardization – Consistent extraction and structured outputs across datasets. Extensibility – Modular, open-source design for adding new data sources and features. Hydra for configuration management (YAML + CLI overrides) Same Config Same Data API (cdsapi, dwdapi) Open Servers Cloud Database Earth Engine climate datasets obs., reanalyses, projections SciWI n 3. Advanced Scientific Workflows Standardized Handling Open and Modular code Open-Source (MIT license), Modular Objects with similar function structures Extensible: new data, extreme event definitions added through configuration scripts Access to heterogenous data sources with units handling, format conversions, post processing. Extreme events calculations with extensible “yaml” config for user defined functions. Downscaling through statistical and deep learning techniques. Diffusion and UNET models under development. LLM integration and Web dashboard development (Future Work) Hydra: config.yaml, indices.yaml, datasets.yaml Inputs: point (tuple), box (dict), or shape (shp or geojson) get_<dataset>_<subset>.py Workflow outputs: Data in multiple formats (<dataset>_<subset>.<format>); CF compliant Metadata Alignment with FAIR Consistent code structure; published as pypi package Netcdf and zarr metadata is CF convention compatible with provenance information Formal Documentation underway DOI and Zenodo release 🡪Customizable extraction, user defined extreme indices handled with yaml files 🡪 Structured naming convention for easier reuse and interoperability. Simulation and Data Science pypi package