Day 75: ECMWF 96-120 Hour Forecast

Posted by on 16 Mar 2010 | 0 comments

Day 75 (a)

As I mentioned last night, big changes are looming on the horizon for people in the southern plains of the United States.  Tonight I’ll examine two different forecast fields for two different forecast times as simulated by the European Center for Medium-Range Weather Forecasting (ECMWF) numerical model, run at 12 UTC this morning (Tuesday).  The first image is valid 12 UTC Saturday (7 AM CDT) and the second image is valid 12 UTC (7 AM CDT) Sunday.

One of the things that should have been conveyed in last night’s post is that there exists a potential for snow across portions of the southern plains between Friday night and Sunday morning.  One of the things that a meteorologist will examine when forecasting for snow is the temperature of the 850 mb pressure level (roughly 1.5 km above ground level over the ocean).  If this level is above 0C, or 32F, then snow is extremely difficult to get at the surface, however other winter precipitation types (sleet or freezing rain) would be possible if the surface temperature is below 0C, or 32F.  If this level is below 0C, or 32F, then snow is possible, assuming precipitation did fall.

The image above is the forecast 850 mb temperatures for 12 UTC Saturday, followed by the 12 UTC Sunday 850 mb temperatures.  As one can see, temperatures across a good potion of Kansas, and the western 2/3 of Oklahoma, are below 0C (32F) by 12 UTC Saturday (7 AM CDT).  This would mean that if precipitation were to occur, snow would be one of the possible precipitation types about which to worry.  By 12 UTC Sunday (7 AM CDT), the 850 mb temperature (shown below) has cooled even further across an even larger area – including much of Arkansas.

Day 75 (b)
Day 75 (c)

This, in and of itself, would certainly suggest that meteorologists would need to be concerned about the potential for snow, should precipitation occur.  However, I haven’t said anything about whether precipitation (and, in particular, snow) would be falling at the two times discussed above.  The plot immediately above and below attempt to address the, “Will is precipitate?” question.  These two images are the 700 mb relative humidity fields – which are being used as a proxy for precipitation fields, since those are not freely available.

Most snowflakes (the big, pretty ones anyways) tend to develop around 700 mb in what is known as the dendritic growth zone, if sufficient moisture can be found at this level.  Based on this, if the relative humidity is high at the 700 mb layer, we can (to a first approximation) assume that snowflakes are certainly possible in the cloud.  If we couple that with the temperatures at 850 mb being below 0C (32F) we can make a first guess forecast as to whether snowflakes are possible to reach the surface.

If we notice the 12 UTC (7 AM CDT) Saturday forecast, we see that the relative humidity is extremely low at 700 mb across central Oklahoma.  This is known as the dry slot, and is a good indicator that precipitation would not be occurring at the surface.  However, notice the high relative humidity values to the south and west of central Oklahoma at this time.  By 12 UTC (7 AM CDT) Sunday, both of these features have moved east of Oklahoma.  Since I do not have forecast images from 12 UTC Saturday and 12 UTC Sunday, we are left to assume that the positions of the Saturday morning features have translated to their new locations Sunday morning.  (This far into the future, that is a relatively safe approximation.)  If we do this, we can assume that the moist air at 700 mb will traverse the state of Oklahoma, with 850 mb temperatures (first two plots above) below 0C (32F), during the day Saturday.  This would indicate that snow is a possibility (but not a certainty) during the day Saturday throughout central Oklahoma.  It will certainly be interesting to watch this unfold during the next few days…