The traditional art of forecasting3 min read . Updated: 03 Sep 2012, 07:08 PM IST
If IMD or a private forecast provider would be willing to use the data from private stations, then the quality of forecasts could improve
September 2012. A month ago, it seemed as if drought was not just imminent, but a certainty. Today, the all-India monsoon deficit is down to 12%. Many parts of the country are still facing drought. If I were a farmer and had to base my agricultural operations on the forecasts of the Indian Meteorological Department (IMD), I’d be in deep trouble.
We are planning to send a mission to Mars and we can’t ensure that farmers get reliable and timely weather forecasts! The credibility of recent endeavours by private service providers offering forecasts and advisories using mobile phones has been affected by their dependence on IMD’s information. Is there a way forward? Should farmers instead rely on their own traditional knowledge?
June 1992. On a really hot summer day, while Dhanji and I were walking from one dhani (hamlet) to another in the Ranjitpura area in Bikaner district of Rajasthan, Dhanji stopped and asked me to look down and describe what I saw. I could only see lots and lots of sand! I was asked to look more carefully. I squatted and peered closely and finally noticed ants carrying grain. Ants carrying grain out of their homes! My first lesson in traditional weather forecasting. If in the month of Ashad ants carry their food out of their homes, it is likely to rain shortly and the ants are moving their food to safety!
In the desert, if the ker flowers first, there will be a drought. If the khejri flowers first, there will be a zamana (a good harvest). Akha Teej is also the time for farmers to plan their agricultural season. In Kumaun, the festival of Harela is preceded by sowing of seeds of five or seven different crops. Many families still believe that this helps establishing which crops should be sown. Every region of this country has similar customs.
Today, farmers are loath to talk about traditional weather forecasting practices. Education has led to them questioning traditional wisdom based on generations of observations. Traditional wisdom, in their view, does not have a scientific basis and relevance today.
Weather forecasts that are accurate or even reasonably reliable can help inform agricultural operations. When to sow or harvest, or when to spray or apply fertilizer? Events related to weather like a sudden rise in temperature or a prolonged dry or wet spell can also be indicative of pest and disease outbreaks in the context of particular crops. In other words, crop advisories linked to weather forecasts can help farmers plan their crop protection better.
It would be foolish to expect any forecast to be 100% accurate. What one needs to strive for is greater reliability in the forecasts. One could argue in favour of a particular mathematical model or process instead of another. But to begin with, we need to recognize that the basis of these forecasts is poor. There aren’t adequate reliable weather stations across the length and breadth of this country to provide IMD with real time data to inform the forecasts.
There has been a huge spurt in private automatic weather stations in the country. A significant number of these stations have GSM chips embedded in them to relay the data periodically. If IMD or a private weather forecast provider would be willing to use the data from these private weather stations to add to its own data, then the quality of the forecasts could improve. Localized weather forecasts could be offered in return for the data thus sourced. Increasing the number of sources of data is the first step in improving the forecasts.
I’ll hedge my bets till then. A mix of Internet-based forecasts and traditional methods of weather forecasting for me.
V.K. Madhavan has worked in the not-for profit sector for two decades and spent 15 years living and working in deserts and hills. He’s still on the fringe asking questions and looking for answers. He will write every fortnight. Comments are welcome at firstname.lastname@example.org