Meteorology: Folk wisdom
The Economist, November 24th - 30th, 2001.

CAN traditional rules of thumb provide accurate weather forecasts?
Researchers in Junagadh, India, are trying to find out. Most farmers in the region grow one crop of peanuts or castor per year. In a wet year, peanuts give the best returns, but if the rains are poor, the more drought-tolerant castor is a better bet. In April and May, before the monsoon comes, farmers decide what to plant, buy the seed, prepare the soil and hope for the best. An accurate forecast would be extremely helpful.

Little wonder, then, that observant farmers have devised traditional ways to predict the monsoon's timing and character. One such rule of thumb involves the blooming of the Cassia fistula tree, which is common on roadsides in southern Gujarat. According to an old saying which has been documented as far back as the 8th century, the monsoon begins 45 days after C. fistula's flowering peak. Since 1996, Purshottambhai Kanani, an agronomist at Gujarat Agricultural University, has been collecting data to test this rule. He records the flowering dates of trees all over the university's campus and plots a distribution to work out when the flowering peak occurs. While not perfect, C. fistula has so far done an admirable job of predicting
whether the monsoon will come early or late (see chart).

Similarly, with help from local farmers, Dr Kanani has been investigating a local belief regarding the direction of the wind on
the day of Holi, a Hindu festival in spring. The wind direction at certain times on Holi is supposed to indicate the strength of the
monsoon that year. Wind from the north or west suggests a good monsoon, whereas wind from the east indicates drought. Each year before Holi, Dr Kanani sends out postcards to over 400 farmers in Junagadh and neighbouring districts. The farmers note the wind direction at the specified times, and then send the postcards back.

In years of average and above-average monsoons (1994, 1997, 1998 and 2001), the wind on Holi tended to come from the north and west. In the drier years of 1995 and 1996 the majority of farmers reported wind from the east (Dr Kanani did not conduct the study in 1999 and 2000). As with the C. fistula results, the predictions are not especially precise, but the trend is right.

Dr Kanani first became interested in traditional methods in 1990, when an old saying attributed to a tenth-century sage named Bhadli-that a storm on a particular day meant the monsoon would come 72 days later-proved strikingly correct. This prompted Dr Kanani to collect other rules from old texts in Gujarati and Sanskrit.

Not all of his colleagues approve. Damaru Sahu, a meteorologist at Gujarat Agricultural University and a researcher for India's
director-general of meteorology, says that traditional methods are "OK as a hobby". But, he goes on, they cannot be relied upon, and "may not be applicable to this modern age." Yet Dr Sahu concedes that meteorological science has failed to provide a useful alternative to traditional methods. For the past 13 years, he notes, the director-general for meteorology has predicted "normal monsoon" for the country. Every year, the average rainfall over the whole country is calculated, and this prediction is proved correct. But it is no use at all to farmers who want to know what will happen in their region.

Dr Kanani hopes that his research will put traditional methods on a proper scientific footing. He and his colleagues have even set up a sort of peer-review forum for traditional meteorology. Each spring, he hosts a conference for 100 local traditional forecasters, each of whom presents a monsoon prediction with supporting evidence-the behaviour of a species of bird, strong flowering in a certain plant, or the prevailing wind direction that season. Dr Kanani records these predictions and publishes them in the local press.

He has also started a non-governmental organisation, the Varsha Vigyan Mandal, or Rain Science Association, which has over 400 members. Its vice-president, Dhansukh Shah, is a scientist at the National Directorate of Meteorology in Pune. By involving such mainstream meteorologists as Dr Shah in his work, Dr Kanani hopes to bring his unusual research to the attention of national institutions. They could provide the funding for larger studies that could generate results sufficiently robust to be published in peer-reviewed science journals.