Ag Weather Forum
Subtle Differences in Weather Models Lead to Big Differences for Next Week
Model forecasts have been more difficult to trust over the last month. Repeatedly, the extended range of both the American GFS and European ECMWF models have shown variability, not just with their day-to-day temperature and precipitation forecasts over the U.S., but also with their extended range indicators for patterns that meteorologists look toward to give us hints about the coming weather and how important they will be. It can be frustrating with so much variability and not a lot of clarity.
Models are usually better at picking up on the large-scale, upper-level features of the atmosphere than the finer details of a forecast, as we would expect, which is why meteorologists try to deduce what will occur at the surface by what we see in the upper levels. Ridges of high pressure usually mean heat and dryness, while troughs are cooler, with wetness in between the two -- typically wetter when the trough is to your west than when the trough is to your east. But these are only general implications we get from the upper-level pattern. What actually happens at the surface is more complicated than that. Meteorologists often look to models to help discern the specifics more than just a hot-cold-wet-dry forecast and these specifics are the things that matter to those looking at a forecast.
So, when models do a poor job or are inconsistent, it leads to a low-confidence scenario. That is usual for the period beyond the next 14 days, but rare within the next 10 days. Of course, models do have differences within their next 10 days of forecasts. It's why meteorologists look at a lot of them. But many times, the differences are subtle such as temperatures being in the middle 80s as opposed to the upper 80s or lower 90s. Or precipitation being forecast to be light and spotty as opposed to being more organized. What is unusual is when the upper-level forecasts look nearly identical but give polar opposite prescriptions for temperatures and precipitation. It does happen, but it makes it frustrating.
Such has been the case the last few weeks and another scenario like that has unfolded for next week, Aug. 18-24. Models are prescribing an upper-level ridge to expand from Texas northward into the Canadian Prairies this weekend and be a stable feature for next week. A trough is forecast to be situated off the West Coast with another across the Eastern U.S. The ensemble forecasts of the American GFS (GEFS) and European ECMWF (EPS) are very similar, but there are some subtle differences. The GEFS has the ridge more into the Plains while the EPS has it closer to the Rockies. That subtle difference allows for big changes down the line. With the ridge in the Rockies, a trough in the East will be allowed to pull down a stronger cold front from Canada and spread cooler air throughout the Corn Belt and really all areas east of the Rockies outside of Texas. On the other hand, the GEFS presentation allows for much higher temperatures in the Plains and keeps the cooler air only in the Midwest and Northeast, a much more stable block and less chaotic.
But the more interesting piece to the puzzle may be the operational runs of both models. Being only one of the ensemble members of both the GEFS and the EPS, the GFS and ECMWF runs show somewhat opposite of their ensemble forecasts. The GFS looks more like the EPS and the ECMWF looks more like the GEFS.
So how do you make a forecast with so much uncertainty? Some meteorologists use model blends to get the big picture details. And that works fine for an extended forecast. But when the timing is so close in the future, as is the current 6- to 10-day forecast for Monday-Friday, Aug. 19-23, meteorologists sometimes have to go with their gut and risk being very wrong. Communication about these differences therefore is paramount for end users to understand the risks involved in the forecast.
For my part, I believe the GEFS/ECMWF look will be the most likely, with a warm or hot Western Corn Belt and a milder or cooler Eastern Corn Belt. The front that drops down may bring some showers with it, but precipitation will be much tamer next week than this week across most of the Corn Belt. The Northern Plains and Canadian Prairies will probably see increased chances for precipitation, though.
To find more weather conditions and your local forecast from DTN, head over to https://www.dtnpf.com/…
John Baranick can be reached at john.baranick@dtn.com
(c) Copyright 2024 DTN, LLC. All rights reserved.