By Sukumaran Narayanan
Most of the Godavari basin floods after 10 to 11 days of rain, while the semi-arid Mancherial sub-basin floods after about two, according to a new study in Hydrology and Earth System Sciences.
The study, titled Role of hydrometeorological variables and catchment area to flood generation over the monsoonal climate region of India, has Saran Aadhar of the Indian Institute of Technology (IIT) Jodhpur as the corresponding author and Efrat Morin of the Hebrew University of Jerusalem and Utkarsh Gupta of IIT Jodhpur, as co-authors.
The Godavari river, India’s second-largest river basin after the Ganges, rises at Triambakeshwar in Nasik district, Maharashtra, and runs 1,465 km, the study says. Its basin drains about 312,800 square kilometers across Maharashtra, Telangana, Andhra Pradesh, Chhattisgarh, Odisha and Karnataka.
The researchers combined 53 years of daily rainfall data, from 1967 to 2019, with a computer model of the Godavari river basin. The rainfall came from the India Meteorological Department’s gridded record, built from about 6,995 rain gauges.
The Variable Infiltration Capacity (VIC) model used is a computer program that tracks how rain soaks into soil, runs off the land and flows down to river gauges. The team checked the VIC model against daily river flows measured by the Central Water Commission at five stations between 1967 and 2014.
The authors counted the Godavari floods from the VIC model’s simulated river flows for 1967 to 2019, because measured records after 2014 had too many gaps. A flood was any daily flow above the 95th percentile, meaning the highest 5% of all daily flows. Peaks less than 10 days apart counted as one flood.
The study defined a high-intensity rain day as one above the 99th percentile, the wettest 1% of days. The team then tested which rain totals, over windows of different lengths, best matched each flood peak.
The study split the Godavari into five sub-basins, each named after the gauging station at its outlet, and classes Mancherial as semi-arid and the other four as semi-humid. The list below gives each sub-basin’s catchment area and flood count.
- Pathagudam: catchment about 35,625 km2, semi-humid, 88 floods in 1967–2019.
- Mancherial: catchment about 84,375 km2, semi-arid, 116 floods.
- Tekra: catchment about 90,625 km2, semi-humid, 107 floods.
- Perur: catchment about 226,250 km2, semi-humid, 105 floods.
- Polavaram: catchment about 257,500 km2, semi-humid, 104 floods.
From the study’s figures, the five stations logged 520 floods in total, or about 9.8 a year combined. Mancherial recorded about 32% more floods than Pathagudam. The Polavaram catchment covers about 82% of the whole Godavari basin area.
What did the study find about rain and Godavari floods?
Floods in Tekra, Pathagudam, Perur and Polavaram matched about 10 to 11 days of accumulated rain and 7 to 8 days of runoff. By The Fourth Plate’s calculation, the semi-humid rain window is five times longer than the Mancherial window.
Around 40% to 50% of floods in those four semi-humid sub-basins came from several high-intensity rain days while about 30% to 40% of floods in Mancherial, Perur and Polavaram followed long spells of low-intensity rain.
The authors found that season also shaped Godavari floods. Heavy rain in June and July lead to less high river flows because the soil is largely dry then. Heavy rain in August and September more often became a flood, as soils were already wet. Mancherial also showed a higher ratio in October, which the authors link to the northeast monsoon.
The central and downstream parts of the Godavari basin contribute most to floods with the Tekra sub-basin having the highest runoff coefficient – the share of rain that becomes runoff – under both wet and dry soil conditions.
The study left out human activity on rivers, including reservoirs, irrigation, groundwater pumping and urban growth. The authors write that the work “does not consider the extensive role of human activities on water resources”. The study used only five river gauges, which limits detail at the scale of smaller catchments. The authors also state that the VIC model “underestimates excess runoff and infiltration estimations during intense monsoon events”. The flood counts rest on simulated flows for the whole 1967 to 2019 period.
The authors call for rainfall forecasts covering 10 to 11 days over the central and downstream Godavari basin, to guide flood and reservoir management. They also recommend real-time soil moisture monitoring in the Tekra sub-basin and suggest natural flood management, such as tree planting and changes in land use, to cut runoff in Tekra.
For farms and towns on the lower Godavari in Telangana and Andhra Pradesh, the study implies that flood warnings must track rain totals across 10 or more days. A wet August or September raises the flood risk from each heavy storm. Indian flood records have long been studied this way, as in P. C. Mahalanobis’s flood statistics for Bengal and Orissa.
Sources: Aadhar, S., Morin, E. and Gupta, U. (2026). Role of hydrometeorological variables and catchment area to flood generation over the monsoonal climate region of India. Hydrology and Earth System Sciences, 30(21), 6747–6761. https://doi.org/10.5194/hess-30-6747-2026. Published under a Creative Commons Attribution 4.0 (CC BY 4.0) licence. No press release was found; all findings are from the paper.
