As the third month of the monsoon draws to a close, the India Meteorological Department continues to stand by its prediction that this season’s rainfall will be 98% of the long-period average.

The department predicted in April that India as a whole would receive 99% of its average monsoon rainfall. It revised the prediction in June, and provided a region-wise break-up: North West India would get 96% of the long-period average rainfall, Central India 100%, the southern peninsula 99% and North East India 96%. The prediction’s margin of error was set at 8%.

Long-period average refers to the average rainfall between 1951 and 2000, which the Met department uses to determine how far from normal rainfall is at present.

The department does not yet predict rainfall for sub-divisions although it has promised block level predictions by 2019. “We do not have the capability for issuing forecast for individual substations,” said Mrityunjaya Mohapatra, scientist at the National Weather Forecasting Centre. “We are working on improving our systems and going in for a dynamic modelling system. We still have a month and a few days before we can verify our forecasts for the season.”

Numbers game

As of August, it is increasingly evident that should these numbers actually hold true by the end of the monsoon season, it will only be because of averages.

Take North West India, which was supposed to get 96% of its average rainfall. As of August 24, West Rajasthan had received 57% more rainfall than normal while East Rajasthan got 2% less than normal. The average for the state would then be 27% more than normal.

This might be outside the department’s predicted 8% margin of error, but taking into account that western Uttar Pradesh has seen 36% less rainfall than normal, and Punjab, Haryana, Chandigarh, Delhi, Himachal Pradesh and Uttarakhand have all received less than normal rain as well, the predicted number might just average out.

Look closely at monsoon figures and this is evident. Poor rainfall in a few districts of some states gets masked when tallying the state average. State figures get subsumed in regional data and regional in national. Figures also get covered in time.

An article in Down to Earth examined weekly monsoon data for this season. It found:

 “So far about 400 weekly observations from 36 meteorological sub-divisions have been listed. Less than a fourth of these fall under the normal category, and yet rainfall over the country as a whole is normal.”  

KJ Ramesh, director general of the India Meteorological Department, told Business Standard on Monday that one way of testing whether there has been adequate rainfall was to look at kharif sowing data. Seasonal data, released by the Ministry of Agriculture on August 18, shows that area sown has declined by 0.8% compared to the previous year.

Again, these figures only mask regional differences, the Hindu Business Line reported. Sown area in Karnataka, for instance, declined by 27% over the last year.

Low reservoir levels

Rainfall or sowing data are not the only indicators for understanding how India will tackle the dry months that follow the monsoon. Reservoirs are an important source of water in the several parts of the country, particularly in peninsular India.

Here, a grimmer picture emerges. As of the week ending August 17, according to the latest data available with the Central Water Commission, 35 of India’s 91 reservoirs had less than 20% of their live storage, that is, water that can be used productively for irrigation, drinking and generating electricity.

The level in three reservoirs in Maharashtra and one shared by Andhra Pradesh and Telangana is at 0%. While these reservoirs are not necessarily dried out, a level of 0% indicates the water in the dam cannot be disbursed.

It is not unusual for reservoirs to regularly have low live water storage, particularly if upstream users consume excess water, or if that region is prone to low rainfall. That is why the Central Water Commission also provides data showing the 10-year average of live storage.

Of the 91 reservoirs that the commission monitors, 62 have live storage between 30% and 100% less than the 10-year average. As many as 16 have 0% live storage, or worse. (Another anomaly here is Tripura’s Gumti dam whose live storage level is a whopping 341.7% more than the average. It had no live storage in 2016.)

The regional distribution of the reservoirs with low storage too shows that it is not prudent to rely on monsoon rainfall figures alone to understand how India can manage water more efficiently.

The most acute water scarcity is in Western India, which accounts for 20 of the 62 reservoirs with lower live storage. Most of these are in Marathwada and Vidarbha, regions that have long been prone to water scarcity. Eastern Gujarat too has low levels.

Tamil Nadu may seem to have recovered from low rainfall in July, but its reservoirs still have low levels. Dams in the west, along the Kerala border, as well as in central Tamil Nadu are nearly running dry. These areas have suffered at least three consecutive seasons of poor rainfall. Meanwhile, the live storage levels in lakes nears Chennai continue to hover around 90% less than on the same day last year, according to data from the Chennai Metropolitan Water Supply and Sewerage Board.

So, while the Met department’s map shows Tamil Nadu receiving normal or near-normal rain, its reservoirs evidently do not reflect this.