Guide to making accurate snowfall forecasts:

[FLguy]

This tutorial (which is my presonal guide to snowfall forecasting) is partly intended to help others understand the multitude of different factors which go into making an accurate snowfall forecast … and to answer a question in a previous thread in regard to snow/liquid ratios.

CAUTION...THIS IS A METEOROLOGICALLY INVOLVED DISCUSSION.

In order to make an accurate snowfall forecast one must understand and be able to operationally apply the following factors to their forecast:

1) Ability to make an assessment of which model(s) has the best storm track.

2) One must be able to understand whether the synoptic pattern is set-up for a major/large scale snow event or a modest/localized event.

3) Understanding of low-level model temperature biases. This is especially important with the GFS which quite frequently has a problem with keeping low level temperature fields TOO WARM.

4) The models handling of the mesoscale structure.

ACCURATE SNOWFALL FORECASTING INVOLVES THE FOLLOWING PROCEDURES:

1) Correct analysis of the surface and 850mb low pressure tracks depicted by the model or models of choice.

2) Deciding where the RA/SN/mix line will set-up based on the track of the surface and 850mb low pressure centers. This goes hand in hand with rule number 1.

3) Making a determination of whether the low level temperature fields WILL or WILL NOT support snow accumulations providing that the mid-level thermal profiles support snow (this can be determined by proper analysis of vertical soundings).

4) Decide what the snow to liquid ratio will be. Quite often snow to liquid ratios can start off high at the beginning of an event...and end with a low snow to liquid ratio. Or the reverse where the ratio starts low and increases by the end of the event.

PHYSICAL FEATURES CORRESPONDING TO SNOW ACCUMULATIONS:

The physical features which decide how much snow falls in any given location involve the relationship between the vertical transport of moisture...cloud physics which can affect precipitation processes and precipitation efficiency. How expansive is the snow-producing cloud shield and are additional snow producing clouds located upstream to continue the precipitation.

FORECASTING PRECIPITATION TYPE:

Precipitation type is dependent on the vertical thermal profile of the atmosphere over a given location. The forecaster can make an assessment of the vertical temperature profile over a specific location be examining atmospheric soundings over the area in question. One can use 1000-500 thicknesses…1000-850 and 850-700 partial thicknesses but neither of those two methods can resolve thin warm layers aloft which can affect precipitation type even in spite of those two parameters favoring or implying all snow.

Atmospheric Soundings are HANDS DOWN the best method by which to make the MOST SKILLFUL assessment of the deep layer thermal profile over any given location to determine precipitation type.

The depth of the below freezing layer can determine what type of precipitation one will experience. In the case where the entire sounding is below freezing…snow will be the favored precipitation mode. If there are warm layers in the mid-levels…with a relatively deep layer of sub-freezing temperatures below it…sleet is favored. If the mid-levels are solidly above freezing with a VERY shallow layer of cold air trapped close to the surface…FRZRA is likely to be the preferred precipitation type.

If the WBZ (wet-bulb zero) and freezing levels are very high and the surface is ABOVE freezing the precipitation type will be rain.

Now aside from that…one must decide whether or not there will be changes in precipitation type during the event in question.

To do that we must look at:

1) Evaportational processes:

If specific layers within the sounding are close to or slightly above freezing in an un-saturated environment ESPECIALLY IF A DEEP LAYER OF THE ATMOSPHERE IS ISOTHERMALLY NEAR 0C …as moisture moves into the unsaturated air mass…evaporative processes result in the cooling of the layers which are borderline freezing. If the deep layer temperature profile when saturated (wetbulb temperatures below freezing…actual temperatures near 0C) is below freezing…snow will be favored. If the actual temperatures in the mid-levels are TOO WARM that the wetbulb temperatures (Tw) are solidly ABOVE freezing…however below the above freezing layer…a below freezing layer exists once the atmosphere becomes saturated…depending on the depth of that cold layer…either FRZRA or sleet will be favored.

Here are the basic parameters for precipitation type based on wetbulb temps:

-- If Tw (or the wetbulb temperature) exceeds +3C snow melts resulting in rain or FRZRA.

-- Tw is LESS THAN +1C…the snow partially melts but will likely re-freeze.

-- Tw is +1-3C…the snow partially melts but will eventually re-freeze. Sleet favored in this situation however depending on the depth of the warm layer…FRZRA must be considered.

If the lower cold layer is less than -10C and frozen nuclei are in abundance while spending a long duration of time in the cold layer…snow or sleet is the favored precipitation type. When the layer is WARMER than -8C …droplets are super-cooled if the snow is melted. This scenario favors FRZRA.

The temperature of the cold layer is far more important than its depth…though both should be well analyzed by the forecaster.

SNOWFALL INTENSITY FORECASTING:

Snowfall rates are dependent on the correspondence between the growth rate of a single crystal and the number of crystals per unit volume. Normally crystal growth rates are maxed out at around -15C. When cloud top temps are greater than -25C the concentration of ice particles is sufficient to use all available moisture within the stratiform and orographic clouds.

The number of ice crystals per unit can be increased when more fragile crystals such as dendrites or needles fracture…ice splintering during riming…and or during the fragmentation of large super-cooled droplets when freezing.

Flake size is dependent largely on aggregation (many ice crystals make-up the same snowflake). The aggregation process is maximized when temperatures are about 0C (32F)

The physical characteristics of snowflakes are important because the favored crystal type may affect snow/liquid ratios. This is also known as the Fluff factor. Un-rimed dendrite and plate ice crystals have the lacy structures which promote the highest snow/liquid ratios and best overall precipitation efficiency (in other words you get the most accumulation for your QPF…lol). When snowfall growth is forced by riming (due to the presence of unfrozen liquid in the clouds) snow/liquid ratios are LOWER. 10:1 or less usually.

The size and the composition of the snowflake may also decide how it sticks to the ground in the event that surface temperatures are marginal.

Larger flakes will take longer to melt in an environment which is marginal temperature-wise for accumulations.

SNOW/LIQUID RATIOS:

Marginal surface and boundary layer temperatures will tend to reduce snow/liquid ratios and decrease precipitation efficiency leading to less accumulation. A warm layer with temperatures near 0C will keep ratios down as well.

Clouds with a greater number of super cooled water droplets will have lowers snow/liquid ratios. This as crystal growth is dependent more on deposition.

Soundings with a deep isothermal layer near 0C will probably have a snow/liquid ratio between 8:1 and 10:1.

IT SHOULD BE NOTED THAT WHILE COLDER AIR DOES PROMOTE HIGHER SNOW/LIQUID RATIOS. ONE SHOULD BE AWARE THAT GIVEN A VERY COLD ENVIRONMENT THE CRYSTAL TYPE MAY NOT BE FAVORABLE FOR MAXIMIZED RATIOS…DECREASING PRECIPITATION EFFICIENCY AND LESS OVERALL ACCUMULATION.

Nor’ Easters (and system which have a track close to the ocean) will NOT have as high of ratios as say what an Alberta clipper would as there is more liquid present in the cold its self. Decreasing ratios and precipitation efficiency resulting in LESS snow accumulation.

In the eastern part of the country…the lower the 1000-500 thickness value is…the higher the ratio will be.

540 DM thickness ----- 10:1 ratio.
528 DM thickness ----- 17:1 ratio.

The above is just to give you a basic idea… values can vary significantly.

These are short range snowfall and precipitation type forecasting methods to be used in order to make accurate accumulation forecasts.

Cold season convective processes such as CSI (Conditional Symmetric instability) can result in localized higher accumulations under enhanced precipitation bands which cannot be resolved by the models. CSI occurs when the absolute vorticity of the geostrophic flow becomes weakly positive or negative. This results in inertial instability which is mixed out as shear increases. The CSI is the atmospheres response to the disturbance (which of course is the negative absolute vorticity since the earth’s natural vorticity is positive).

CSI can be assessed by examining cross sections of theta-e and momentum surfaces (when the slope of the theta-e surfaces is greater than the momentum surfaces…the environment is considered favorable for the release of CSI). While that is a good method…cross-sections of EPV and frontogenesis can give us a better idea of where the environment may be symmetrically unstable.

Remember also that the environment MUST BE SATURATED in order for CSI to develop. areas of negative EPV located in layers above a sloping frontal boundary normally imply a favorable environment for the release of CSI and development of banding. The forntogenesis can continue to disrupt the environment and result in prolonging the life-cycle of CSI.

Strong vertical motions promoted by negative omega and frontogenesis can lead to heavier QPF…which given favorable snow/liquid ratios can result in heavier accumulations. (Frontogenesis is described by the horizontal tightening of the thermal gradient over a given location)

[Suzi Q]

Great explanation but now I feel like I've started classes again and I still have one more week off!!!! What surprises me is I actually understood about 1/2 of it. Reinforces again how little I know.

[donsutherland1]

FLguy,

Fantastic tutorial. This is great reference material for S2K. I really believe that there should be a folder where such material is categorized and maintained for ready access.

Great job.

[Suzi Q]

I absolutely agree!!! There were times last semester when I was completely befuddled in my first met class and could come on here and voila! There it was explained a little more clearly than what my prof or the book had said. And seeing as I have such a long road to hoe, this kind of explanation and those to follow, will truly help me when I get stuck!

Marva the Muddled

[Stephanie]

I usually copy and paste information like this into an excel spreadsheet for my reference later.

THANK YOU so much for that tutorial Flguy!

[FLguy]

I absolutely agree!!! There were times last semester when I was completely befuddled in my first met class and could come on here and voila! There it was explained a little more clearly than what my prof or the book had said. And seeing as I have such a long road to hoe, this kind of explanation and those to follow, will truly help me when I get stuck!

Marva the Muddled

first of all...thanks everyone for the kind words. and i hope this helps folks understand just what snowfall and precipitation type forecasting involves. even with that said...these are only the basics.

eb...since i dont know how yor professor teaches i cant comment. but i will say this. its been my observation that many different students learn in many different ways...which is why a particular style of teaching works better with some students better than others. in other words which it may have been harder to understand for you ... it may have been easier for others to understand.

given the large numbers of students it can be hard for a professor to isolate best the way each of his students learn so he/she can modify their teaching methods to correspond to that their learning style.

i know the material can be very hard to understand at times. and its only going to get harder. but you need to take what you do know and do understand to decifer the rest of the material. Much of that is self-teaching. since in meteorology everything is tied together. what effects one thing will effect another...and that will then effect sonething else and so on down the line.