If you’re keen on studying the weather, my first piece of advice: don’t start with the objective of becoming a forecaster overnight. An interest in the weather is what you really need. It’s a self-learning process that takes years to perfect. For most, an interest in the monsoon is a seasonal affair; once the rains recede, the interest does too.
First steps
Start by studying local weather patterns and terminologies; a simple Google search should do the trick. I am against learners following apps and blogs, as the information there is tricky. Once you’re familiar with the basics, I recommend learning Nowcasting on imd.gov.in, which involves interpreting radar images based on your understanding of cloud formation, moisture content, etc, to predict weather for the next few hours. It’s a skill that takes months to perfect, but is the safest tool for a beginner. No one is an expert when it comes to tracking weather patterns. Everyone is dependent on technology and it’s the knowledge you’ve acquired, your interpretation skills, and years of patient practising that set you apart.
Get familiarised
For weather forecasts, Numerical Weather Prediction (NWP) is followed worldwide. We have three ranges: short range forecasts for about 10 days (useful for cinema professionals to plan their shoots, developers to alter construction schedules), medium range for 10 days to one month, and the long range monthly model that is useful to track trends. For instance, how the NE monsoon this year will pan out, will it be above or below normal, etc. But the areas it will hit – say Chennai, Puducherry or Cuddalore – can be identified only as the days near. I pass on information about seasonal trends to farming communities across Erode, Tirupur, Coimbatore and Thoothukudi, who in turn notify farmers.
My pattern
For short range forecasting, I use the ECMWF, GFS and UKMET weather models, and each prediction depends on which model picks up a particular weather movement. I track them at night and put out a Facebook post in the morning stating the probability of rains through the day. But I do so only when I am a 100% sure.
Once the rains start, I keep a close track of the radar and continue with updates about the spell’s intensity, duration and the areas it will hit. I also keep in mind what the models have predicted for the day. People are unaware why a certain model shows a cyclone or a depression, unless they have studied it for years. Many, just by looking at an image, predict the worst. This is what I try to avoid. It’s important what information you put out there, especially in Tamil Nadu, as people are psychologically affected due to earlier disasters they have witnessed.
What a computer predicts in the morning is different from what it picks up in the latter part of the day. The downpour we witnessed earlier this month wasn’t predicted – models showed about 30-40 mm rain in the city, but we received over 300 mm, 10 times over.
Radar image of cyclone Vardah that hit Chennai in December 2016
The way you pass on information is crucial. I cover a wide range of topics and make my posts as interactive as possible. Giving out advice is important, too. People here are worried about a flood or cyclone but thunderstorms and lightning are the real killers. I inform people about a possible storm and how to stay safe.
On days when heavy showers are expected – like during cyclone Vardah – I need more time to study the models that run two to four times a day. As every run differs, their performance alters – it can worsen or even get better – which makes our job trickier. Today, a certain model could show heavy rainfall, but in a subsequent run there would be no signs. I will soon start a YouTube channel and take to FB live on important days and take questions from followers.
- imdpune.gov.in
- tropmet.res.in
- imd.gov.in
- timeanddate.com
- Dos: Know your sources, keep a personal record of data, track for a month or even longer to get a hang of weather tracking and be consistent
- Dont’s: Start forecasting immediately, make it a seasonal affair, create panic, spread rumours
Why it’s tough
Tricky TN: It’s the toughest region in the world to forecast the weather for, as it is a very small basin. Unlike other parts of India, between April and the first week of October (before the onset of the monsoon), we get intense thunderstorms. Rainfall patterns are erratic too, as the Northeast monsoon can't be predicted in advance. We get various cyclones that other metros are not familiar with.
India patterns: Weather patterns here are very different from other countries. In the US, if they predict a hurricane, it's an accurate reading. But that’s not the case in India even though we run the same weather models. The models struggle to provide accurate data and keep changing often. Even if you get a leading meteorologist from abroad to study our patterns, he will be thoroughly confused.
How it’s done: On November 2, the UKMET model predicted a cyclone and BBC forecasted 500 mm of rainfall. Since I’ve been tracking these models for years I am aware of what we can expect. I followed the ECMWF and GFS models and there was no cyclone. Both models were interpreting the same data but their resolution differed. BBC isn't wrong always but things pan out differently at times. It is also wise to follow an ensemble model: cumulative of previous individual and multi-modal runs.
Public interest
Since 2015, the number of weather bloggers in the State has shot up. Chennai tops the list, followed by Mumbai, Bengaluru, Kolkata and New Delhi. I started my blog in 2008 and the FB page in 2014, but they became popular during the 2015 floods. It’s taken people time, but there is a lot more curiosity about rainfall patterns today. Many track radar images and ask me for suggestions.
Step up to schools
To make future generations self-sufficient, introducing weather studies in schools is an opportunity waiting to be tapped. I have lectured in over 20-25 schools and colleges in the city, and realised we don't have a single chapter on numerical weather prediction in Geography textbooks. Teaching children early on about our city's climatology will help tremendously.
Data mining
After storm chasing, it is is my next favourite hobby. I closely monitor and record rainfall patterns, tsunamis, earthquakes and solar eclipses. Be it earthquakes in TN over the last 200 years or a list of tsunamis that have hit India over the past 250 years, I have spent weeks gathering data from various sources.
