Comments on: Weather Prediction – How Far Can You Go? https://blogs.mathworks.com/community/2014/07/11/weather-prediction-how-far-can-you-go/?s_tid=feedtopost News from the intersection of MATLAB, Community, and the web. Fri, 24 Oct 2014 21:54:37 +0000 hourly 1 https://wordpress.org/?v=6.2.2 By: Ned Gulley https://blogs.mathworks.com/community/2014/07/11/weather-prediction-how-far-can-you-go/#comment-69063 Fri, 24 Oct 2014 21:54:37 +0000 https://blogs.mathworks.com/community/?p=2762#comment-69063 Hi John:

Thanks for the note. The first number in each column of collected data is the *actual observed high temperature* for that day. This gets flipped around to be the last column of the d2 matrix. The error analysis is showing how the forecasted high for a given day converges on the actual high for that day as you get closer to it.

]]>
By: John M. https://blogs.mathworks.com/community/2014/07/11/weather-prediction-how-far-can-you-go/#comment-69061 Fri, 24 Oct 2014 21:08:15 +0000 https://blogs.mathworks.com/community/?p=2762#comment-69061 Hi Ned,

Very interesting project, I’m actually interested in doing this for my local university. I do have one question about this methodology though.. so correct me if I’m wrong. In your code, you’re pulling the forecasted temperatures for 10 days off of weather.com, and the first entry in your code is the forecasted maximum temperature.. not necessarily the actual maximum temperature for that day. With this in mind, isn’t your error analysis showing the error of forecasting the forecasted maximum temperature, rather than the error of forecasting the actual maximum temperature?

If this is the case, I would expect a more chaotic pattern in the error analysis due to atmospheric variability or local effects. Including in the maximum temperature from the Norwood Memorial Airport (http://w1.weather.gov/obhistory/KOWD.html) or Worcester Regional Airport (http://w1.weather.gov/obhistory/KORH.html) as your validation temperature (i.e., Day 0) would create a more robust error analysis, or so I imagine. What do you think about it??

]]>