Mea culpa on Sri Lanka’s weather forecasting


Posted on June 23, 2017  /  4 Comments

I saw a response to an RTI request from the Department of Meteorology on Twitter and did not adequately check its veracity. As a result, I unfairly described the forecasting capabilities of the Department in at least one occasion at a meeting attended by influential officials and also polluted Twitterspace. I am sorry.

Above was where I made the mistake. My mistake was that I did not check the date on the RTI response that had been posted:

The downpour that caused all the deaths and devastation occurred on the night of the 25th of May. On the 25th of May, the Department of Meteorology issued an amber advisory entitled “Weather advisory for heavy rainfalls valid for next 48 hours” stating that more than 150 mm of rain was likely. This was significantly below the 553 mm that actually fell. But they did use words like “heavy rainfall” and did not give a number as low as 50 mm, as shown in the Groundviews tweet that showed the May 16th advisory.

Even two days before the event, on the 23rd of May, the Department issued a “Special weather advisory for heavy rainfalls and strong winds” where it warned of “heavy falls (more than 100 mm).”

The May 16th advisory shown by Groundviews had no relevance to the question of the adequacy of the forecasts provided by the Met Department, being issued nine days before the event.

This is not the first time I have erred by placing too much faith in the veracity of tweets from non-professionals. But I have no excuses. Mea culpa. Thank you to the Head of Forecasting at the Department who politely corrected me. I am sorry for the mistake and for taking so long to correct it.

4 Comments


  1. Hello,

    Just wanted to respond to this post. We have clearly indicated that the forecast we tweeted was for May 17 in the original tweet. We noticed that the article you posted was dated May 27. The reason we did not respond to correct you as to the dates was because we did an article search for actual rainfall on May 17. We discovered there was still a fairly high discrepancy.

    See below for articles:

    ReliefWeb: http://reliefweb.int/map/sri-lanka/sri-lanka-severe-weather-echo-daily-map-17052016

    CNN: http://edition.cnn.com/2016/05/18/asia/sri-lanka-flood/index.html

    As ReliefWeb shows, there was over 100 mm of rainfall recorded on May 17. Trust this clears up any confusion you might have on this score.

  2. Notice that you’ve made an addition to this post:

    “The May 16th advisory shown by Groundviews had no relevance to the question of the adequacy of the forecasts provided by the Met Department, being issued nine days before the event.” http://webcache.googleusercontent.com/search?q=cache:TOLcZUcz5iYJ:lirneasia.net/2017/06/mea-culpa/+&cd=1&hl=en&ct=clnk&client=safari

    The floods began before May 25 – in fact, there were already landslides and heavy flooding by May 18: https://www.nytimes.com/2016/05/19/world/asia/sri-lanka-landslides-flooding.html The request we filed was not just for May 17, but also included the forecasts on May 23rd – 26th. We did not share all of these as we are still waiting for additional information from the Department of Metereology.

    We would also like to reiterate that it was you who posted the Daily Mirror article without cross-checking the dates. Mistakes do happen, and are understandable, but we do not believe we made a mistake here. Would appreciate a direct response, rather than an edit made to the post without acknowledging our comment.

    1. My post is headed mea culpa. I accept responsibility.

      The events that most people are concerned started on May 25th night. The Met Department is pointing out that they issued a 150mm< severe weather warning on the 23rd. Amantha Perera referred to that warning, not anything days prior. In my first response on June 1, I referred to the 150 mm< warning. It was because of my inattention to the date that then I started referring to 50 mm, instead of 150 mm. My fault.

  3. Thank you for your response. As we said, mistakes happen. What fuelled our initial response is what you note in your post, that somehow the “veracity of tweets from non-professionals” was to blame. There are two entities in question here. The Daily Mirror. And Groundviews. We cannot speak on behalf of the former but do not believe in this instance they were unprofessional in any manner or placed information in the public domain that was inaccurate. We have explained our position on the matter including our RTI applications for more info which the Met Dept isn’t answering in full and delaying. We fail to see how the responsibility of what is an understandable error in comprehension can be in any way placed on what you in this post impugn are inaccurate and unprofessional sources. Thanks again for clarifying your position.