RE: Another example of incorrect upslope snow placement.

Forums (Forecast Builder)

RSS (Opens New Window)

Jeffrey Hovis, modified 8 Years ago.

Another example of incorrect upslope snow placement.

Youngling Posts: 1 Join Date: 9/24/12 Recent Posts

I know that I have posted information about an issue that we are seeing with the placement of upslope snow west of the mountains in the past. I also know that the issue may be caused by the resolution of some of the models that are included in the ForecastBuilder/Superblend data. However, I captured another example of this issue this morning. We have a "clipper" that is expected to impact our region Thursday into Friday of this week. So we will have snow associated with the clipper and then upslope snow on the back side of the system. In the examples that I have attached (StormTotalSnow and SnowAmtRange), the heaviest snow amounts are located west of the mountains. In fact, higher snowfall amounts are actually located across some of our lowland counties instead of the west-facing slopes which usually receive the heaviest snows from northwest upslope flow. Because this occurs with every upslope snow that we have seen so far this year, this is much more than a "Target of Opportunity". We were discussing whether it would be scientifically sound if we write a tool that would simply display the upslope snow amounts a certain distance to the east of where they appear in the ForecastBuilder/SuperBlend data. The forecasters would simply have to determine which part of the storm was synoptic and which was upslope and run the tool. Thoughts?

Andy Just, modified 8 Years ago.

RE: Another example of incorrect upslope snow placement.

Padawan Posts: 99 Join Date: 6/2/15 Recent Posts

Jeffrey,

I think what you show here is like something perfect to send to model developers. First, though, I would check the individual components of SuperBlend to see what's dragging it down. A quick way to do that is to look at PoP and QPF from the CONSRaw and CONSMOS. This will give you an idea if it's model data or MOS data. From there then you could dig through the individual model components of both data sets. I know it's a bit time consuming.

From what I have looked at with the NationalBlend, I think it will perform better for you. We are working on some improvements to PoP for the SuperBlend as well, to yield higher values, though I'm not sure if that helps for your situation as what you are dealing with is more location.

The idea of a tool is not a bad idea to counteract the model bias, but see if you can get it to work on QPF. Keep the end goal of SnowAmt as a derived field.

David Barjenbruch, modified 8 Years ago.

RE: Another example of incorrect upslope snow placement.

Youngling Posts: 21 Join Date: 9/2/14 Recent Posts

Hi Jeffrey,

We've dealt with this issue of models displacing QPF for a long time in the complex terrain of Colorado. To start with a better initialization, the smart inits are adjusting QPF up over the terrain, and down in the valleys. This helps to get at least get the right picture in place before further refinements are made. In addition to the simple up/down terrain, we can also take advantage of a locally run orographic precipitation model, which accounts for many parameters such as wind direction, speed, moisture depth (mixing ratio, cloud water), lapse rate, etc. While there is no hydrometeor drift in this simple 2 dimensional model, it can help with diagnosing orographically produced precipitation and snowfall.

I've attached a couple examples of the QPF smart inits, and you'll note the adjustments made (0.03 additional per 1000 feet over the model terrain, and 0.02 reduction per 1000 feet below model terrain). You can note big differences, roughly doubling the mountain top QPF versus valley QPF in our upcoming precipitation event. Obviously, those can be adjusted, but we've found those to work pretty well here in our normally drier climate. A lot of Western Region sites also make use of PRISM data and that has also shown to be extremely beneficial. We hope some of this can get adapted in the NBM.

RE: Another example of incorrect upslope snow placement. - Forums - ForecastBuilder

Forums (Forecast Builder)