← Back to project catalogue
GP-CV-189WXEMCivilOpen for request

Watershed health remote-sensing dashboard

A geospatial analysis dashboard for tracking land cover, vegetation, surface water, erosion proxies, and rainfall patterns within selected Indian watersheds.

  • Nuxt
  • Vue
  • FastAPI
  • PostgreSQL
  • PostGIS
  • Python
  • Rasterio
  • GeoPandas
  • Docker

Project definition

Problem statement

Watershed condition changes through land-cover conversion, vegetation loss, surface-water variation, erosion, and rainfall extremes, but the available datasets differ in date, resolution, cloud cover, and measurement meaning.

The engineering problem is to align public geospatial data, calculate transparent indicators, measure change consistently, and distinguish remotely sensed evidence from conclusions that require field inspection.

Project objectives

  • Delineate or import selected watershed and sub-watershed boundaries.
  • Prepare cloud-filtered satellite, terrain, rainfall, land-cover, and surface-water layers.
  • Calculate documented vegetation, water, slope, drainage, land-change, and erosion-risk indicators.
  • Compare indicators across seasons and years using consistent spatial units.
  • Rank sub-watersheds for field review using transparent weights and sensitivity analysis.

System design

System modules

01

Geospatial data manager

Catalogues source, date, resolution, projection, quality, licence, and processing history for each layer.

02

Watershed processor

Prepares elevation, flow direction, drainage, boundaries, slope, and sub-watershed units.

03

Indicator engine

Calculates vegetation, surface-water, land-cover change, rainfall, and selected erosion proxies.

04

Change analyser

Compares aligned periods and records changed area, direction, confidence, and possible data artefacts.

05

Map dashboard

Displays layers, time series, sub-watershed scores, uncertainty, and exportable field-review maps.

Methodology

System workflow

  1. 01
    Select watershed

    The user opens a prepared boundary, period, seasons, and available public data layers.

  2. 02
    Prepare layers

    Data is reprojected, clipped, masked, resampled, and checked for coverage and cloud limitations.

  3. 03
    Calculate indicators

    Each sub-watershed receives documented terrain, vegetation, water, land, and rainfall measures.

  4. 04
    Analyse change

    Comparable periods are evaluated and likely artefacts are separated from stable changes.

  5. 05
    Prioritise review

    Weights are applied transparently and sensitivity analysis shows how field priorities change.

Demonstration scenario

A watershed is compared across the same season in two years. The dashboard shows vegetation reduction and a surface-water change in two sub-watersheds, then overlays slope and rainfall indicators. Changing the prioritisation weights reveals which field-review areas remain consistently high priority.

Engineering

Technical architecture

Web application
Nuxt and Vue for map layers, timeline controls, indicator charts, comparison, and exports.
Geospatial API
FastAPI for layer metadata, raster jobs, vector analysis, indicators, rankings, and tiles.
Spatial data layer
PostgreSQL and PostGIS for watersheds, sub-watersheds, drainage, vector layers, indicators, and provenance.
Raster pipeline
Python, Rasterio, and GeoPandas for terrain, imagery, rainfall, masking, zonal statistics, and change analysis.
Reproducibility
Versioned source manifests, fixed projections, processing configurations, quality masks, and prepared validation areas.

Testing

Evaluation

Evaluation measures

  • Boundary and drainage agreement with prepared reference data
  • Indicator correctness against manually calculated sample areas
  • Land and water change accuracy on prepared validation regions
  • Sensitivity to cloud masks, spatial resolution, season, and threshold settings
  • Ranking stability under alternative indicator weights
  • Processing time across raster resolution, period count, and watershed size

System boundaries

  • The indicators describe selected remotely sensed and modelled conditions and do not constitute a complete watershed assessment.
  • Cloud, resolution, classification, boundary, and seasonal uncertainty are shown with each result.
  • The system does not infer a legal cause, land ownership, or responsibility for a detected change.
  • Field surveys and qualified civil, hydrology, and environmental review remain necessary.

Included

  1. 01Watershed map and time-series analysis interface
  2. 02Image preprocessing, indicator, change, and prioritisation modules
  3. 03Land, vegetation, water, slope, and rainfall visualisations
  4. 04Prepared public geospatial data, scenarios, and results
  5. 05Complete source code in a private GitHub repository
  6. 06Complete project documentation with synopsis, abstract, methodology, watershed maps, analysis results, screenshots, and conclusion
  7. 07Setup and usage guide

Project record

No buyer information is collected on this page.

Permanent project ID
GP-CV-189WXEM
Catalogued
21 Aug 2026
Completed
Pending
Verified
Pending
Demonstration
Added when ready

Handover

After purchase

  1. 01
    Payment is confirmed

    The project is marked unavailable and cannot be purchased again.

  2. 02
    Repository access is granted

    The buyer's submitted GitHub account receives access to the private repository.

  3. 03
    The purchase record is delivered

    The certification sheet is prepared from the reviewed buyer details and sent privately by email.