Skip to Navigation Skip to content

Regional and Mesoscale Meteorology Branch

Search the RAMMB website

Precipitation Type Forecasting


Dan Baumgardt



Archived Training





The Precipitation Type Forecasting teletraining session by Dan Baumgardt covers the following topics:

  • Review of Cloud Microphysics (Ice Nucleation)
  • Operational Application of Microphysics
    • Hydrometeor Altering Environments
    • Top-Down Approach
    • Operational Forecast Problems
    • Applications, More Tools

The two objectives of the session are:

  1. To trace a hydrometeor trajectory from its source to the surface to determine its final precipitation type and
  2. To understand why a precipitation type is occurring because it will help you forecast it.

Training Session Options

NOAA/NWS students – to begin the training, use the web-based videoYouTube video, or audio playback options below (if present for this session). Certificates of completion for NOAA/NWS employees can be obtained by accessing the session via the Commerce Learn Center

  1. Audio playback (recommended for low-bandwidth users) – This is an audio playback version in the form of a downloadable VISITview and can be taken at anytime.

    Create a directory to download the audio playback file (49 MB) from the following link:

    After extracting the files into that directory click on either the visitplay.bat or visitauto.bat file to start the lesson. If both files are present, use visitauto.bat

References/Additional Links

  • Talking points are available for this lesson and may be printed out to easily review the session in detail at any time.

Post session exercise:

  • Near the end of the training, I will show a cross-section. I will discuss it a bit but not completely. This would seem to be a very good exercise for the trainees to complete AFTER the VISIT session to reinforce the topics. You can grab a copy of this from (print with landscape). The solution is under

List of references:

  • Pruppacher, H.R., and J.D. Klett: Microphysics of Clouds and Precipitation. D. Reidel Publishing Company, Boston, 714 pp. See page 241+ as well as page 9
  • Rodgers, R.R., and M.K. Yau: A Short Course in Cloud Physics. Pergammon Press, New York, 293 pp. – See page 155+ also page 161
  • Schichtel, M., 1988: Specification of Precipitation Type in Oklahoma Winter Storms. University of Oklahoma. Masters Thesis. 131 pp.
  • Pobanz, B.M., J.D. Marwitz, and M. K. Politovich, 1994: Conditions Associated with Large-Drop Regions. Journal of Applied Meteorology, 33, pp 1366-1372. In this study, 29 cases were used which evaluated cloud-top temperature for large liquid drops. Only 14% of the liquid water drop cases had temperatures at or below -12.0°C. About 31% of the cases had temperatures of -10°C or colder.
  • Stewart, R.E. 1985: Precipitation Types in Winter Storms. Pure Appl. Geophys., 123, 597-609.
  • Czys, R.R., R.W. Scott, K.C. Tang, R.W. Przybylinski, and M.E. Sabones: A Physically Based, Nondimensional Parameters for Discriminating between Locations of Freezing Rain and Ice Pellets, Weather and Forecasting, Volume 11 1996, pp 591-597.
  • Zerr, R. J.: Freezing Rain: An Observational and Theoretical Study. Journal of Applied Met., Dec 1997, pp. 1647-1661.
  • Penn, S., 1957: The prediction of snow versus rain. Forecasting Guide No. 2, U.S. Weather Bureau, 29 pp.
  • Lussky, G., 1995: Physical Rational Associated with Various Minneapolis, Minnesota Heavy Snow Checklist Parameters. NOAA/NWS 4th National Winter Weather Workshop, Kansas City, MO. pp. 33-1 to 33-7.

Try it yourself, use the Precipitation Type Applet that makes use of these concepts to see what type of precipitation would fall.

Talking Points

Slide numberTalking points
Slide numberTalking points


Title slide


Topics to be covered



The Top-Down Approach is a technique to assist in forecasting precipitation type

Knowledge of why a precipitation type is occuring now will increase your chances of correctly forecasting precipitation type further into the future.


Frequency of freezing precipitation in hours per year

Note two maxima over the plains and northeast, however it can occur in MANY areas of the US.


Understanding ice nucleation

Note: In a cloud saturated layer, if the temperature is warmer than -4 °C there will not be any ice forming in that layer unless it is introduced from another source (i.e. another cloud over the top seeds the clouds below with ice).


Particles that are ice nuclei

Kaolinite is the most common ice nuclei


IN’s can’t activate unless they are -9 °C or colder


Sounding from Jackson, Mississippi

Coldest temperature in the saturated layer is -8 °C, therefore it is most likely supercooled liquid water. Ice can’t be initiated within this layer (unless it is advected in from another source). Precip type here would most likely be freezing drizzle.


Important temperatures for ice nucleation


Data collected from flying through clouds. 

Concentrate on lines (1) and (2)

At -8 °C they ran into all supercooled liquid clouds about 55% of the time while 45% of the clouds had SOME ice.

At -14 °C they ran into all supercooled liquid clouds about 28% of the time while 72% of the clouds had SOME ice.

BIG difference between -8 °C and -14 °C


These numbers fit well with activation numbers we saw earlier


Figure from Mike Schichtel

Observed precipitation type from sounding information

Minimum cloud temperature is within the saturated layer

Predominance of snow cases (red line) are on the colder side of -10 C

Predominane of freezing rain cases (green line) are on the warmer side of -10 °C

13Important temperatures when you’re trying to diagnose if there is going to be ice introduced into a layer.

-12 °C is the key temperature. If minimum temperature in saturated layer reaches -12 °C there is a pretty good likelihood that ice will be introduced into the cloud

14Omaha sounding from 1973, key on -12 °C

Ice is introduced because a good portion of the saturated layer is colder than -12 °C

15Now look at other hydrometeor altering environments and what else that ice can run into as it moves towards the surface
16Elevated warm layers above freezing

Look at warm layer depth and warm layer maximum temperature. Hydrometeor size is also important but is difficult to diagnose operationally therefore will not be considered here.

171973 Omaha sounding

Consider the blue triangle

Read off warm layer maximum temperature quite easily

18How does the warm layer maximum temperature relate to the depth of the warm layer (melting layer).

Don’t worry about what precipitation type was observed in the graph of warm layer depth vs warm layer maximum temperature. Concentrate on the family or population of points in this diagram. Relationship is linear, as depth of the warm layer increases, the maximum warm layer temperature increases proportionately.

19Warm layer depth is represented by the blue line on the sounding

Maximum warm layer temperature is represented by the light green line

Go through the sequence of soundings, as the warm layer depth increases, do does the maximum warm layer temperature. 

20Warm layer temperature will be considered from this point onward since it is easier to assess operationally.
21Table shows the expected precipitation type given this warm layer

For the 3rd column, collision and coalesence produces the precipitation since there is no ice present

2nd column

< 1 °C, ice doesn’t have a cold enough temperature to melt the snow

1 to 3 °C, is the transition area with sleet by 3 °C

22Warm layer between 1 and 3 °C is highlighted as the transition area (mix)
23Penn, S., 1957: The prediction of snow versus rain. Forecasting Guide No. 2, U.S. Weather Bureau, 29 pp.

Consider synoptic scale warm advection on the order of 10 °C / 12 hours at 850mb being fairly significant. The environmental cooling refered to here is an order of magnitude greater than synoptic scale warm advection changing the sounding significantly in the matter of a few hours.

24LBF sounding

Note very small warm layer (1 °C)

Wet bulb temp is quite a bit lower than that, therefore any warm layer would be removed in time.

If warm layer is unsaturated it will cool quite rapidly to the wet bulb temperature.

25Seeder – Feeder mechanism

Ice is introduced into a supercooled liquid cloud below which glaciates supercooled liquid cloud (ice grows by the Bergeron process).

26Theoretical study. Ice particles were dropped into a sounding enviornment in a model. Parcel trajectories were traced on how far they made it vertically before they totally sublimed.

Right side is the sounding with the temperature profile (red) and dewpoint profile (orange).

Left side: Particles were dropped from around 375mb, 1’s at the end of them are the lower ice concentrations, 4’s are the higher ice concentrations.

Particles 1A and 1B (smaller ice concentrations) sublimed within the nearly saturated layer. Higher ice crystal concentrations made it into the unsaturated layer below 500mb then lasted about 1 km. Some of the sublimation from the ice in the higher ice concentrations took place in the saturated layer, then fell into the unsaturated layer. Lower bound of about 3500 feet where the ice crystals can seed into liquid layer below.

3500 feet is the minimum distance and 5000 feet as the maximum distance between 2 layers that you can count on introducing ice all the way through an unsaturated layer.

27Tools to help you determine hydrometeor altering environments

LAPS soundings find the cloud base, IR 10.7 um satellite imagery for cloud top temperature, upstream raobs, radar VWP (see slide 28), IR 3.9 um satellite imagery to determine ice or water clouds

28Case study: WFO MKX used the seeder-feeder idea to forecast a change in precipitation type.

At 01:04 UTC, saturated cloud layer to 7000 feet, above is cloud layer 11,000 feet and up, cloud base cannot be determined exactly, freezing drizzle was observed at MKX at this time.

With time the gates get lower and at 02:13 UTC the two gates come together meaning ice falling into a supercooled liquid layer, at this time there was a changeover to snow at MKX.

29Keep in mind temperature of terrestrial objects

Surface cold layer deeper than 2500 feet and colder than -6 °C comes from Zerr study

30Start at the top and ask “is ice going to be introduced into this environment?”

Draw in -12 °C line on sounding, in this case the region colder than -12 °C is unsaturated therefore ice is not expected here.

Middle portion look for warm layer or unsaturated layer. There is a warm layer, however if ice is not introduced into the warm layer it doesn’t matter what the maximum warm layer temperature is, it’s going to be a freezing drizzle outcome if surface layer is supercooled.

Bottom surface cold layer. Saturated layer 0 to -6 °C meaning no ice, supercooled liquid. Freezing rain would be the result at the surface assuming no cloud is overhead seeding the lower cloud.

31Summary of Top-Down approach
  OKC sounding

Assess this sounding. What kind of precipitation type do you expect?

Saturated at upper levels (in the region of -12 °C or colder) so ice falls into the warm layer.

Warm layer max temperature is about -2 °C (mix). After falling into the cold layer at lower levels it will freeze as ice pellets. Ice pellets were observed, could have a slight mix of snow in there as well. 

33DFW sounding

Assess this sounding. What kind of precipitation type do you expect?

Saturated at upper levels (in the region of -12 °C or colder) so ice falls into the warm layer. Warm layer maximum temperature is about 5 °C which would melt into liquid then fall into cold surface layer. Mostly freezing rain, possibly with ice pellets mixed in would be the result.

34Plan view using same case as in previous two soundings in DFW and OKC.

Draw in 1 and 3 °C lines and see how the area in between these isotherms corresponds to surface observations. It fits pretty well (mix of sleet or snow in between 1 and 3 °C with snow north of the 1 °C line and freezing rain south of the 3 °C line – but there are a FEW outliers

35Volume browser function in AWIPS

Choose temperature and coordinate would be all mb.

Go through the loop to see temperatures at various pressure levels, use this to find the warm layer maximum temperature and investigate it in plan view to diagnose precip type.

36Assess this sounding. What kind of precipitation type do you expect?

Saturated at upper levels (in the region of -12 °C or colder) so ice falls into the warm layer. Warm layer temperature is around 1 °C so there is a possibility of mix, surface cold layer is around -10 °C therefore expect mostly snow. 

Verification: about 80% snow, with about 20% sleet which comes from that warm layer maximum temperature around 1 °C.


37Assess this sounding. What kind of precipitation type do you expect?

Unsaturated around -12 °C therefore no ice moving into the warm layer. Warm layer is unimportant in this case. Supercooled liquid drops possibly coalescing  falling and freezing on contact at the groundand. Freezing drizzle was observed

38Assess this sounding. What kind of precipitation type do you expect?

Elevated convection casues ice to fall into the warm layer which is 1 °C. Cold air is dragged from above as well so the warm layer either stays constant or decreases in temperature. Bottom layer is below freezing therefore snow is the result.

39Snow and ice pellets are the result (from slide 38)
40IR 10.7 um satellite imagery

Note occluded front from eastern Iowa to northern Illinois, warm surge of air heading towards Lake Michigan associated with higher cloud tops/more ascent. Over the LaCrosse, WI CWA we see elevated convection (colder cloud tops) move over an area of freezing rain causing the precip type to change to snow.

41IR 3.9 um satellite imagery

Cloud tops that are water in white, cloud tops containing ice are dark. White area in western Wisconsin changes to darker colors. This represents elevated convection moving over this area changing the precipitation type from freezing rain to snow.

42Radar loop that corresponds to the 2 previous satellite loops.

Note stable snowbands over MN/northwest Wisconsin. Convective bands are shown in white. Area around radar becomes “red” (see radar enhancements) as ice is introduced into the environment.

43Assess this sounding. What kind of precipitation type do you expect?

Saturated at upper levels (in the region of -12 °C or colder) so ice falls into the warm layer. Warm layer maximum temperature is about 1 °C so there may be a mix of ice pellets, then the particles fall into colder layer near the surface. Result is mostly snow with perhaps some ice pellets mixed in. Also keep in mind vertical motion (determine if the vertical motion occuring in that layer – not easy to asses – will warm layer warm or cool?) being generated in “close” cases like these.


4410.7 um IR satellite image. Enhancement: blues show temperatures warmer than -12 °C, whites -12 to -15 °C, pink -15 to -20 °C, -20 °C and below is black

Black over most of western WI, however the observations show different types of precip type, LAPS sounding from the previous slide would suggest snow and perhaps a little sleet.

After calling around, it was indeed snowing at most locations. A little sleet did mix in briefly which caused these observations. 

Make sure you call around to observers if you have automated obs in your CWA.

453.5 hours later, same imagery as slide 44.

At ? ob (gray shading) it had just stopped snowing and it changed to freezing drizzle. The call was made at the time of the sat image.

46LAPS sfc ptype icons under SFC2D icon on AWIPS.

LAPS tries to diagnose precip type at sfc. Icons are present only where radar data is present.

In this case LAPS showed snow in that snowband.

47LAPS snow accumulator. Storm total precip.

Snow accumulation off because of bright banding, but had orientation of snowband perfectly in this case.

48Plan view to visualize how the warm layer changes spatially

Draw in isotherms where you would expect your warm layer to change. 3 °C near the freezing rain area, all snow would be around 1 °C.

49Cross section from ND to IA which was depicted in plan view on slide 48

Fill in the boxes with the precip type expected across this region (shown on a map in slide 48)

Answer sheet can be found off the student guide web page

50Flow chart of Top-Down approach to forecasting precipitation type
51Other problems to consider when forecasting precipitation type
52continued from slide 51
53Example of a LAPS sounding in error on the warm layer temperature, remember to check the accuracy of LAPS soundings before diagnosing precipitation type
54Plan view from previous case to assist in verifying LAPS sounding
This course is Basic

There are no prerequisites


Dan Bikos


Page Contact

Bernie Connell


Unless otherwise noted, all content on the CIRA RAMMB: VISIT, SHyMet and VLab webpages are released under a Creative Commons Attribution 3.0 License.