How often you have found yourself, smirking while listening or reading to weather forecasts. More often than not, isn’t it? And the general lament is why can’t the met guys get it right? Then there are some who take the opportunity to tell the world, “Oh when I was in Britain they predicted the weather an hour in advance and it was so accurate you could plan your day according to it”. The people listening to them almost always have that “wow-am-I-impressed?” look on their faces.
However, a little deeper understanding of weather forecast will tell you the fault lies not in the science or the technical and intellectual capability of our scientists but in our approach towards communication. Let’s see how?
Gathering Inputs: The Backbone of Weather Forecast
Let’s begin with how the climate and weather data is collected. Earlier we had synoptic observatories. It meant that there was one observatory within 250 square kilometres. Then the Indian Meteorological Department (IMD) decided to gather even more granular data by moving to meso-scale (intermediate) observatories which meant an observatory every 50 kilometre radius. Coupled with the infrastructure on the ground a string of new satellites offered a lot of information that was not available earlier. Traditionally coastal areas were more prone to extreme weather events like cyclones and heavy rains. So as a priority the IMD set up a chain of radars along the coast where almost every inch is now under the watchful eyes of the radar network put up by the department. Satellites offer the initial information of a gathering storm or a cyclone from 600 to 700 kilometres away in the Ocean and the moment it’s 300 kilometres away from the coasts the radars take over to amp its trajectory – Its movements, speed and direction. Then the information is fed into pre-existing models at the IMD office and the predictions are made.
This is one of the reasons that the number of deaths in cyclones has drastically reduced over the years. While the 1999 cyclone in Orissa took a toll of 10000 the toll during cyclone Ockhi that devastated a large part of eastern coasts in India in 2017 was at 245. Given the density of population and the little time there was to evacuate the entire population to safer places it was a huge improvement in pre-emptive safety measures initiated due to timely prediction.
As the times are changing so are the challenges. Now climate variability and climate change both have far reaching impacts on our resources. Himalayas are one geographical feature that may suffer due to extreme climate change. As Himalayas are home to all our perennial rivers, glaciers and ice-caps they need to be studied in depth. The problem is, mountains are very different from the planes and the coasts. While the latter have flat surfaces offering a congruity of air and temperature spread the former has a very complex environment due to its special geography.
In the mountains temperature will vary at the foothill or in the valley, at the middle of the mountain and then at the top where the airflow is absolutely unencumbered. It also depends on which side of the mountain you are measuring the temperature. The windward side and the leeward side will have vastly different readings. As the Himalayas have extreme variations in alleviation within districts and blocks and also in just a 100 kilometre width (as the crow flies) it is impossible to decide how many observatories are enough to gather information from every nook and corner of the mountain range and is it possible to establish that many observatories. Take the case of Dehradun district in Uttarakhand, you have the city of Rishikesh which is 372 metres above sea level, while the same district has Chakrata which is at 2118 metres above mean sea level. With this kind of variation even a meso-scale saturation of observatories may not give an accurate picture of the entire weather pattern.
Weather forecasting from the shortest term called now-casting to seasonal forecasts. Now-casting is an American term where the forecasts are made for a few hours ranging from half an hour to three hours. Then comes the short range forecasts that have a time horizon of 48 hours. This is followed by medium range forecast which is applicable for five to seven days. Extended rain forecast is for four weeks and this is followed by seasonal and long range forecast (three months or more).
Typically seasonal, long range predictions involve a larger geography and thousands of inputs from all around the world. It is here that mostly all the world meteorological departments bite the dust. They are on the ball sometimes while missing the mark by a wide margin in other cases. Take the case of extreme winter we are experiencing right now. Almost all the met departments around the world predicted last year that this winters will be a tame affair in the northern hemisphere. But as we saw they all ended up having egg on their face because none could anticipate polar vortex letting loose its icy slice.
Similarly, the larger the geographical area, the more the chances of predictions going awry. Take the case of Delhi. Many times met department predicts rain and people sitting in Mayur Vihar or in Dwarka or in Noida see no rains at all. But within the met department their forecast is completely right as there was rain in Connaught place. The logic is if the forecast comes true within a fifty kilometre radius it would be deemed as correct.
The problem here is that Delhi has expanded exponentially over the last 100 years when the first observatories were set up. From a small city with a population of five lakhs it has grown to a metropolis of close 20 million people. The area has expanded too from few kilometres of dense settlements to even and dense spread across 60 kilometres or more. However, the perceptions haven’t changed and the people in the entire NCR (National Capital Region) think they are part of Delhi and if there is no rain in their part of the city or sub-city the prediction is futile.
While the long range forecasts over a large geographical areas are the most difficult to predict, the easiest one is now-casting. This forecasting system is based on information gathered from radar and has to take into account variables which are far more localised and easy to factor in. The weather forecast for the next few hours is very useful in deciding whether you should saddle yourself with an umbrella, organise an outdoor party, have a cricket or tennis match or organise a political meet.
This is where the US and British weather departments score over the Indian Met department. Over the years they have built up a system where the now-casting information is disseminated through radio, television, sms and mobile app. IMD too has the same ability to predict single day forecast with equal accuracy but the outreach to the people is not as comprehensive.
A sms service has been initiated but a lot more can be done. The mobile phone companies can be asked to preset their handsets launched in India to IMD’s day forecasts app displayed on their screen as a default setting, similarly TV’s can run the scroll as was successfully experimented by the regional weather department in Dehradun. Radio FM stations, websites can also chip with information dissemination.
So in future during drawing room conversations if someone berates our long term weather forecasts, know that we are as good as the others. And when the same people wax eloquent about the real time forecasts in the US and Britain, rueing its absence here, know that this is a communication outreach issue not an indictment of scientific abilities of our weathermen.
the #1 Itinerary on Five things I learned in … indiadynamic on LET’S TALK SERIES: Why online… Dhiraj Singh on LET’S TALK SERIES: Why online… Clean Ganga: Where e… on Clean Ganga: Where experts fai… indiadynamic on Incremental changes in agricul…
- June 2019
- May 2019
- April 2019
- March 2019
- February 2019
- January 2019
- July 2018
- June 2017
- May 2017
- May 2016
- December 2015
- November 2015
- July 2015
- June 2015
- October 2014
- July 2014
- June 2014
- May 2014
- April 2014
- October 2013
- May 2013
- December 2012
- November 2012
- October 2012
- July 2012
- June 2012
- December 2011
- November 2011
- October 2011
- October 2010
- June 2010
- May 2010