Shsg Falls Posted March 4, 2012 Share Posted March 4, 2012 I'm curious how the forecasts at The Weather Channel (or a point forecast after entering a zip-code on weather.com) is created? Does anyone know? I tried a Google search but didn't find anything reliable, and the few weather forecast papers regarding TWC I came across dealt more with verification. 1. Does TWC use a grid based system similar to the NWS? 2. What numerical model data is used? Any special TWC products/numerical output used? 3. Is there any human intervention? 4. How often do their forecasts update? Thanks ahead for your replies. Link to comment Share on other sites More sharing options...
Isopycnic Posted March 5, 2012 Share Posted March 5, 2012 I'm curious how the forecasts at The Weather Channel (or a point forecast after entering a zip-code on weather.com) is created? Does anyone know? I tried a Google search but didn't find anything reliable, and the few weather forecast papers regarding TWC I came across dealt more with verification. 1. Does TWC use a grid based system similar to the NWS? 2. What numerical model data is used? Any special TWC products/numerical output used? 3. Is there any human intervention? 4. How often do their forecasts update? Thanks ahead for your replies. I'm pretty sure it involves random guesswork and pulling stuff out of nether-regions. Link to comment Share on other sites More sharing options...
Analog96 Posted March 5, 2012 Share Posted March 5, 2012 I'm curious how the forecasts at The Weather Channel (or a point forecast after entering a zip-code on weather.com) is created? Does anyone know? I tried a Google search but didn't find anything reliable, and the few weather forecast papers regarding TWC I came across dealt more with verification. 1. Does TWC use a grid based system similar to the NWS? 2. What numerical model data is used? Any special TWC products/numerical output used? 3. Is there any human intervention? 4. How often do their forecasts update? Thanks ahead for your replies. I can't answer for TWC, but it might not be too different from what goes on at Accuwx. Over there, there are four map discussions per day. Initially, the database is populated with (I believe) mostly raw output from the GFS model. Forecasters then circle areas to tweak apparent weather and temperatures. Usually, the shorter term is just adjustments to previous (already human-edited) forecasts. The new Day 7 product sometimes requires more attention, since it's usually just the 168 hour GFS. Link to comment Share on other sites More sharing options...
Ed Lizard Posted March 5, 2012 Share Posted March 5, 2012 I have noticed, since TWC replaced NWS zones with their own forecasts during the 'Local on the 8" segments, I have gained new respect for NWS HGX, as the TWC local forecasts are obviously inferior to NWS products. Link to comment Share on other sites More sharing options...
Mr Bob Posted March 5, 2012 Share Posted March 5, 2012 I'm curious how the forecasts at The Weather Channel (or a point forecast after entering a zip-code on weather.com) is created? Does anyone know? I tried a Google search but didn't find anything reliable, and the few weather forecast papers regarding TWC I came across dealt more with verification. 1. Does TWC use a grid based system similar to the NWS? 2. What numerical model data is used? Any special TWC products/numerical output used? 3. Is there any human intervention? 4. How often do their forecasts update? Thanks ahead for your replies. I am not sure how WSI and TWC are structured after they were bought out from Landmark but they are still tied together at the hip, I believe...and their forecast processes go hand in hand. Dicast was brought in the late 90s. It has evovled to whatever it is called today and most private vendors that supply temperature forecasts to a variety of sources use a similar process...basically a very large ensemble taking in all model data they can get their hands on. There is a staff of VERY experienced forecasters that are looking at the output and post processing where they need to. It has been many years since I worked there so I do not know the specifics anymore... Link to comment Share on other sites More sharing options...
Shsg Falls Posted March 6, 2012 Author Share Posted March 6, 2012 I have noticed, since TWC replaced NWS zones with their own forecasts during the 'Local on the 8" segments, I have gained new respect for NWS HGX, as the TWC local forecasts are obviously inferior to NWS products. I'm glad you brought this up, and this is one of the reasons why I asked the original question about how TWC does their forecasts. I've looked on "Forecast Advisor" and it often shows the NWS forecasts aren't as accurate compared to TWC, but can this really be true? For example: The similarity of verification between TWC vs. Intellicast is almost identical in this example. Link to comment Share on other sites More sharing options...
Rainshadow Posted March 7, 2012 Share Posted March 7, 2012 I'm glad you brought this up, and this is one of the reasons why I asked the original question about how TWC does their forecasts. I've looked on "Forecast Advisor" and it often shows the NWS forecasts aren't as accurate compared to TWC, but can this really be true? For example: The similarity of verification between TWC vs. Intellicast is almost identical in this example. No. There used to be a person from this company on the Eastern Board, don't know if he is still here. The 2011 verification stats for the first six forecast periods (covering three days) three point forecast locations within New York City (Central Park, LaGuardia Airport and JFK Airport) indicated that percentage of maximum temperature forecasts within three degrees was 70.2% and the percentage of minimum temperature forecasts within three degrees was 73.7%. The percentage of correct "pop" forecasts was 86.4%. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.