Why mcdaStudio
Decisions about where to recharge an aquifer, protect groundwater or farm land depend on many factors at once: geology, water availability, land use, regulations and the views of the people affected. Spatial multi-criteria decision analysis (GIS-MCDA) brings these factors together on one map, in a way that can be explained and repeated.
mcdaStudio is the successor of INOWAS T05, the GIS-MCDA tool of the INOWAS platform for managed aquifer recharge (MAR). It replaces the earlier client with a new codebase, a modern interface and server-side raster processing in Python, and keeps the T05 methodology after Rahman et al. (2012).
Pairwise comparison uses the same method as T05, with column normalisation and the T05 random index table, so weights calculated in both tools can be compared directly.
What is new
- An explicit reference grid. Coordinate system, extent and cell size are set at the start; every layer is aligned to it, and the grid can be changed later.
- Vector inputs. Polygons are rasterised, lines and points become distance rasters, and slope is derived from a DEM, all on the server.
- Scenarios and sensitivity. Weightings are stored as named scenarios that can be compared side by side, and the one-at-a-time sensitivity analysis is implemented.
- Accounts, sharing and teaching. Projects have owners, editors and viewers; course templates give every student an independent copy.
- Application templates. One workflow for MAR site suitability, aquifer vulnerability, flood risk, groundwater potential, land suitability and custom assessments.
- Data provenance and confidence. Each layer records its source, licence and a 1–5 confidence rating, summarised for every result.
- Candidate areas and public links. Screening regions for field investigation, and read-only links that show frozen results to stakeholders without an account.
How it is built
The browser handles display and interaction; Python services perform all raster processing. The methodology lives in mcda_core, a Python package independent of the interface and covered by unit tests, with the examples of the T05 documentation as reference cases. Rasters are stored as Cloud-Optimized GeoTIFF; results are exported in open formats (GeoTIFF, ASCII grid, CSV, GeoJSON, GeoPackage) with QGIS styles.
Because mcda_core is independent of the interface, every calculation step is tested on its own and checked against the T05 documentation examples.
| Part | Technology |
|---|---|
| Interface | React, MapLibre GL JS |
| Service | FastAPI, PostgreSQL with PostGIS |
| Processing | Python with rasterio and GDAL, background jobs |
| Map tiles | TiTiler |
| Accounts | Keycloak |
Status
The specification for version 1 and the computational core are complete. The tool is in development and testing at TU Dresden and is not yet open to the public. It will be hosted on TU Dresden infrastructure, where all data are stored.
mcdaStudio is still being tested. Calculations and results may change between versions, so please do not base real decisions on them yet.
Who we are
mcdaStudio is developed by the INOWAS research group at the Institute of Groundwater Management, TU Dresden. Project lead: Dr. Catalin Stefan. See Contact.