Gujarat Surface Water Quality Dataset

This dataset contains surface water quality monitoring records for the state of Gujarat, India, compiled from the Central Pollution Control Board (CPCB) National Water Quality Monitoring Programme (NWMP). The raw data was obtained from the Government of India's open data portal (data.gov.in) under the "Agency-wise Surface Water Quality" catalog.

The dataset covers all 22 districts of Gujarat for which water quality data is available: Ahmedabad, Anand, Banaskantha, Bharuch, Bhavnagar, Daman, Dohad, Gandhinagar, Jamnagar, Junagadh, Kachchh, Kheda, Mahesana, Narmada, Navsari, Patan, Rajkot, Sabarkantha, Surat, Surendranagar, Vadodara, and Valsad. After merging district-wise files, the combined dataset contains 6,816 samples and includes physico-chemical and microbiological parameters measured at surface water monitoring stations across the state.

Files included:

gujarat_full.csv — Merged raw dataset containing all 99 features from the original CPCB monitoring data, including metadata (station code, district, latitude, longitude, date) and all measured water quality parameters.

gujarat_paper.csv — Processed dataset with 12 selected features and a binary target variable. The selected features are Dissolved Oxygen, pH, Biochemical Oxygen Demand (BOD), Fecal Coliform, Total Coliform, Conductivity, Nitrate, Turbidity, Fluoride, Ammonia, Temperature, and Chemical Oxygen Demand (COD). The target variable Water_Safety is binary (0 = Safe, 1 = Unsafe), engineered using CPCB Class C water quality criteria with the following thresholds: Dissolved Oxygen < 4 mg/L, pH outside the range 6.5–8.5, BOD > 3 mg/L, Fecal Coliform > 500 MPN/100ml, or Total Coliform > 5000 MPN/100ml. A sample is labelled Unsafe if any of these thresholds are violated.

Intended use:

This dataset is intended for academic research in machine learning, environmental data science, and water quality assessment. It can be used for binary classification, feature analysis, time-series studies, and the development of predictive models for surface water safety in Indian river systems.

Source attribution:

Original raw data: Central Pollution Control Board (CPCB), Government of India, National Water Quality Monitoring Programme. Available at: https://data.gov.in