3

Part of this assignment is to list the main issues faced by the forecasting group during the week, as discerned from the discussion threads. Below are the issues faced each day of the forecasting period, along with thoughts and observations from the group. I've followed that up with the actual occurrences and some comments about my own forecast in relation to the threads.

A table with the three MOS model forecasts in comparison to the norms and actuals points out where MOS busted on temperatures and winds. I also included my forecast for each day on this table. From this table you can see the challenges that presented themselves during this week - the winds vs. MOS predictions, the precipitation amounts for Days 1 and 2, and the high and low temperatures for Days 1 and 2.

For the first day of our forecast week, discussion began with the timing of the front's arrival being the main forecast challenge. This would affect high and low temperatures due to when cloud cover could be expected to move in, as well as precipitation chances. All agreed that the system was weak and not impressive, and seemed moisture-starved. Discussion hinged around whether to forecast very little precipitation, or none, and whether any of the precipitation that did fall would occur before the end of this forecast period. It was thought at first that the low temperature would occur early in the forecast period before cloud cover moved in. However, this was also when the winds were forecast to be the highest, dying down during the progression of the day, and winds would keep the low temperature up. On the other end of this spectrum, it was thought that perhaps clouds would clear out by the end of the forecast period, and with the lighter winds this would allow evaporational cooling to drop the low temperature considerably. The group finally determined that a low temperature lower than MOS would be appropriate. For the high temperature, the group felt that higher than MOS would be more accurate due to the WAA ahead of the front. Which brought back the challenge of the timing of the front - would it pass early in the period or toward the end.

As it turned out, there were persistent clouds early in the forecast period that moved in earlier than expected, with winds up to 16 knots, which kept the low for Monday night a bit higher than MOS. However, the clouds cleared by 12Z allowing the temperature to drop to 32°F before beginning to climb with solar heating. Clouds moved back in at 18Z and a period of calm winds occurred at 20Z-21Z, allowing the high temperature to reach only 41°F before the winds shifted to the south with the passage of the warm front. On the meteogram below you can see this cloud cover and wind shift, as well as the light precipitation that occurred between 23Z 10/31 and the end of the forecast period 06Z 11/1. Precipitation totaled .02, corresponding with much of the discussion about what amount to forecast assuming precipitation fell at all. I'm happy to say my forecast was for .02 inches! (see Table of MOS-Actuals)

This meteogram for 06z-06z 10/31-11/1 shows the clouds clearing by 12Z 10/31 and a low of 32°F, then clouds moved back in by 18Z-19Z 10/31, winds calmed and a high of 41°F was reached, then the wind shifted to the south-southeast and precipitation occurred by 23Z 10/31. Image from the University of Wyoming site.

On the second day of our forecast week, the amount and timing of precipitation started out being our biggest challenge, but was soon overshadowed by the bigger problem of cloud cover. All agreed the precipitation would occur early in the period, and the minimum temperature would occur toward the end after skies had cleared and cold air advection behind the front had moved in. There was much discussion regarding clouds clearing briefly then reappearing after the passage of the front. It was thought that they would, which set the stage for a high temperature forecast lower than MOS. Winds weren't discussed in this thread, which would have been of benefit since I predicted them to have died down closer to 7 knots for this day.

The cold front followed the warm front and came through with westerly winds and clearing at 12Z. The timing of this frontal passage was well progged for between 12Z-15Z. Winds went up to 17 knots and were quite gusty from 15Z-19Z, and skies remained clear for the rest of the period, allowing the high to climb to 49°F with solar heating and weak cold air advection. This is one of those classic examples of solar heating outweighing cold air advection behind a frontal boundary. The group's consensus that precipitation would occur early in the period was correct and, though moisture-starved, the total amount came in at .03 inches. I'd forecast .02 again. (see Table of MOS-Actuals)

This meteogram for 06z-06z 11/1-11/2 shows the south winds and cloudy conditions through 11Z, with rain at 06Z-07Z 11/1, then clearing skies by 12Z with a wind shift to the west, winds pick up, and temperatures begin to climb by 13Z 11/1. Image from the University of Wyoming site.

On our third forecasting day, precipitation again started out as our main challenge. The model guidance was contradictory and the group waited for the 15Z SREF ensemble to come out to get a better feel for it. This precipitation was related to a low that was forecast to develop over the Atlantic states and move quickly up the Eastern Seaboard. The location, timing and track of this low was the main question, and as the day progressed, it appeared the low would track east of Caribou. It was pointed out that the 700 mb relative humidity dropped in close proximity to Caribou during the time forecast for precipitation chances, therefore dry conditions would most likely prevail. Winds were thought to become light, but once again came out higher than forecast. High and low temperatures were weighed against CAA being offset by solar radiation, cloud cover vs. clear skies, light winds and mixing. The general consensus was that MOS was a bit high on the high temperature (due to cloud cover and CAA) and on the low temperature (due to clearing skies and light winds).

After all the guidance and SREF weighed in, the actual amount of precipitation came in at a trace at 12Z. The low temperature reached 29°F before clouds moved in briefly from 08Z-12Z, when skies cleared again. The high reached 43°F at 19Z with partly cloudy skies, and the clouds remained intermittent through 04Z. The clear skies from 05Z-06Z 11/3 allowed the temperature to drop to 27°F at the end of the forecast period. Unfortunately, I'd forecast the precipitation from the 12Z SREF ensemble, since the models were so disparate, and it was way too much. If I'd waited until the 15Z run, when precipitation came in at 0.01-0.10, I might've gone lower due to the fact that the low was more east of Caribou and so were it's precipitation chances. (see Table of MOS-Actuals)

This meteogram for 06z-06z 11/2-11/3 shows clouds moving in by 07Z-08Z where the low dipped to 29°F, winds picking back up from the southwest-south at 09Z, skies clearing again by 13Z with winds more westerly, the high climbs to 43°F at 19Z, and there is intermittent cloud cover at 23Z 11/2, 02Z 11/3, and 04Z 11/3. Image from the University of Wyoming site.

On the fourth forecasting day, we had fewer challenges than the beginning of the week. At least there was no precipitation to worry about this time. MOS temperature predictions were fairly close to actuals, and the front had passed leaving the ridge of high pressure in place. The group believed that MOS had a handle on high and low temperatures this time, but might still be underestimating cloud cover. MOS also was higher on its wind prediction, though still a bit low, when it had been notoriously low on maximum wind speeds throughout the period. A wind pattern emerged for Caribou that was brought out in the discussion thread: each time the wind speed dropped, Caribou was under the direct influence of a low pressure system, and once the low passed the winds picked back up again.

ETA-MOS verified for the high temperature of 39°F, and NGM-MOS was only one degree off (21°F) of the low temperature of 22°F. The winds came in at 17 knots, which was still 5 knots higher than GFS- and ETA-MOS. Skies were clear at the beginning of the forecast period, then began to cloud up briefly at 12Z, cleared somewhat at 15Z, and became cloudy with a shift to northwest winds at 17Z through 23Z. The low temperature occurred at 11Z under the clear skies just prior to clouding up briefly, and the high temperature occurred during the cloudy and windy conditions between 16Z-18Z. I went with a MOS mid-range for the low temperature, and forecast 22°F, so that worked out well! Also, I went above MOS on the winds with 15 knots, which was only 2 knots below actuals. (see Table of MOS-Actuals )

This meteogram for 06z-06z 11/3-11/4 shows the clear skies and light southwest winds up to 11Z where the low dips to 22°F, then some high clouds move in briefly, skies mostly clear at 15Z-16Z, but clouds move back in and winds become northwesterly by 16Z, with the high temperature of 37°F reached at 18Z 11/3. Image from the University of Wyoming site.

Most of the tools used by all forecasters in the group were HPC maps, HPC QPF, MOS, eWall progs, eWall SREF and NCEP SREF, the local AFD, and forecast soundings. A couple of the class members use Bufkit frequently, and I also used Tabular, Area/Point, and Zone forecasts from the local NWS site. I wanted to point this out to show that we were all using the same tools but many times came to different conclusions regarding what the data were indicating. However, by the end of each discussion thread, the group had for the most part come into consensus on the variables of each day's forecast. This was truly a team effort, and the input from everyone was invaluably beneficial. Many times the threads held conversations full of bouncing off ideas, which helped to form an opinion in the first place. Also, questions were asked and errors were caught that may not have been had we all been on our own.

And what a diverse group - we had morning people who were kind enough to post the local AFD and get the ball rolling, and night owls who stayed up to watch the METARS come in to see how the forecast verified. In between were the posts from different individuals with different styles, different focuses, and using different tools. Together, we were able to make a more informed forecast, which I think is attested to by our standings in the contest.