Understanding river health in India often suffers from a paradox: extensive monitoring exists, yet meaningful insights remain inaccessible. While air quality has benefitted from real-time public dashboards, water quality information still sits inside scattered PDFs, with no platform offering longitudinal, parameter-level analysis. To address this gap, I developed the Yamuna Historical Water Quality Dashboard during my internship with the W2W Research Group at IIT Delhi, under the guidance of Prof. Sovik Das and Anil Dhanda.
The goal was to build not just a visual interface but a scientifically grounded analytical system capable of revealing multi-year, multi-parameter trends with clarity and transparency.

Reconstructing the Data: Building a Usable Historical Dataset

One of the foundational challenges was the raw dataset itself. The Delhi Pollution Control Committee (DPCC) publishes detailed monthly water quality reports, but the data format varies across years and stations.

To stabilise the dataset, I developed a multi-step harmonisation workflow that addressed:

• Station inconsistencies — reconciling location names and merging equivalent monitoring points
• Parameter standardisation — ensuring uniform units and measurement interpretations
• Temporal alignment — mapping data across >10 years into a consistent monthly timeline

This stage alone revealed why there is no existing long-term public dashboard: the cleaning overhead is extremely high.

The dashboard integrates the full WQI pipeline to show how each parameter contributes to WQI at any point in time.

Key features include:

• Real-time WQI computation at a point and changes between two points whenever a user changes year, station, or parameter
• Hover breakdowns showing sub-index values, weightages, and logic
• Parameter-wise trend reconstruction for DO, BOD, COD, pH, coliform, EC, etc.

This design transforms the dashboard from a plot viewer into a scientific inspection instrument.
Researchers can see not just the final score but the full computational chain.

Interactive Visualisation: Turning data Into patterns

The visual layer was engineered to balance academic rigour with accessibility. Instead of static, precomputed charts, every visualisation is generated dynamically.

The interface supports:

• Time-series charts capable of spanning multiple years
• Location-aware mapping to highlight spatial pollution contrasts
• Monthly vs. annual averaging modes
• CSV export functions for research workflows
• An automated insights panel that synthesizes the selected data into interpretable patterns
• WQI Change Analysis tool that allows users to compare any two data points—across time or locations—and instantly compute the difference in their Water Quality Index values

These features convert thousands of table entries into immediately readable environmental narratives—where trends, anomalies, and hotspots become visually obvious.

What the River Actually Reveals

Exploring the dataset through this system uncovers several persistent patterns:

• Seasonal signatures—monsoon flow significantly alters DO and dilution behaviour
• Chronic hotspots—certain stations repeatedly show high BOD and coliform loads
• Inter-annual drift—some parameters slowly improve or deteriorate over multi-year windows
• Discharge influence zones—visualised clearly through spikes near major drains

These insights highlight structural issues in urban wastewater management and the need for coordinated interventions.

---

Positioning for Institutional Use

For organisations such as NMCG, SPCBs, and academic groups, the dashboard offers:

• A reproducible WQI methodology backed by literature
• A unified temporal-spatial view of the Yamuna’s monitored stretches
• Quick anomaly detection for enforcement or field validation
• A transparent communication tool for public awareness

The system functions as both an analytical backend and a public-facing interface — bridging scientific computation with environmental governance.

The architecture generalises well beyond the Yamuna. The plan is to extend this dashboard to Ganga river too.

This project was shaped through continuous mentorship from Prof. Sovik Das and Anil Dhanda, whose guidance helped refine both the environmental and computational dimensions of the work.

I encourage you to share your feedback

https://gandhidaksh.github.io/Yamuna-Historical-Dashboard/yamuna_dashboard.html