BUFKIT: A Software Application Toolkit for Predicting Lake Effect Snow
Edward A. Mahoney* and Thomas A. Niziol
NOAA/National Weather Service Buffalo NY


One of the most significant forecast problems for the Great Lakes Region is the impact of Lake Effect Snow (LES) storms on major metropolitan areas. During the 1995-96 winter season, 28 separate Lake Effect Snow storms generated warning conditions (6 in. or more of snow in a 12-hour period) within the NWSFO Buffalo warning area. One LES storm on December 10, 1995 caused more than 39 inches (100 cm) of snow to fall on the city of Buffalo in just 24 hours (Niziol and Mahoney 1997). To address this significant forecast problem, NWSFO Buffalo meteorologists have developed a highly interactive application toolkit to assist in the prediction of onset, location, and intensity of Lake Effect Snow events building upon techniques developed over the past 30 years (Waldstreicher et al 1996).


A LES band is a mesoscale phenomena forced by many synoptic scale features (Niziol and Levan 1992). During the early 1990s, the NWSFO Buffalo staff were examining nearly a dozen synoptic-level features in the process of producing a LES forecast. Table 1 summarizes some of the more significant features forecasters examine when predicting LES. Up to the early 1990s, forecasters relied upon the standard-level National Weather Service (NWS) AFOS products to predict these features.
LES Forecast 
Contributing Factor(s) 
    Instability Due to Lake-Air Temperature Differences 

    Ice Cover 

    Atmospheric Moisture 

    Low-Level Wind Structure 
    Height of Capping Inversion 

    LES Band Migration 

    Wind Shear Profile 
During the Lake Ontario Winter Storms (LOWS) Project (Reinking et al 1993), participating forecasters developed a checklist where critical LES forecast parameters were weighed and then summed up to provide a "Go - No Go " LES warning decision. At the start of the Lake Effect Snow Project in 1994, the Buffalo office automated this technique as shown in Figure 1. The advantage of this application was that it provided a well-structured approach to addressing the LES forecast problem. However, this approach also had several disadvantages:

- Low temporal resolution. The data were only available at 12- hour intervals. LES band formation, migration and dissipation often occurred between these forecast times. This low temporal resolution lead to a loss of specificity when forecasting LES.

- Low vertical resolution. The decision table made use of only the standard levels available to field offices at that time. Detailed profile forecasts of the atmosphere below 850 mb, critical for LES formation, were not available.

- Limited meteorological insight. The decision table provided little information into what meteorological processes were driving the potential LES event. As in most checklists, the results were viewed as coming from a "black box." The limited insight also resulted in a loss of LES forecasting specificity.

Because of these LES forecasting limitations, the Buffalo staff began working with John Jensenius at the Techniques Development Laboratory (TDL) to develop an experimental data set that provided improved temporal and vertical resolution (Jensenius 1989). For a single forecast grid point, hourly NGM fields were printed out for all 16 sigma levels. The text output of these fields provided the forecasters with the detail to better resolve the low level structure of the atmosphere with the hourly timing. The utility of the data was further enhanced through the use of the graphic plotting of these NGM fields.


During the first year of the Lake Effect Snow Field Project, the Buffalo staff began development of BUFKIT (BUFfalo toolKIT for lake effect snow). The goal was to provide a highly interactive analysis toolkit that allowed forecasters to apply the hourly profile data to the factors critical to LES forecasting as shown in Table 1. The BUFKIT application design constraints included:

- limited processing power. The application was required to run on the office 486 PC.

- limited design time. Software developers were the NWSFO Buffalo meteorologists providing operational forecast and warning support.

- the need for a rapid update cycle. The entire forecast staff contributed to the design of the LES toolkit. During the LES Project, new LES techniques were proposed and tested on almost a weekly basis. The involvement of the forecast staff in BUFKIT's design significantly enhanced their desire to explore the possibilities of this new data set.

- the need for an intuitive user interface. Since the toolkit was used in an operational forecast and warning environment, the application had to be highly interactive and provide almost immediate response to forecaster queries.

To address these design constraints, the NWSFO Buffalo staff designed BUFKIT using the Rapid Application Development (RAD) language Visual Basic. The utility of the RAD approach in quickly developing forecaster applications was demonstrated in late 1995. In November 1995, NCEP began generation of early-eta forecast profiles. Work began immediately to update BUFKIT to take advantage of the early-eta's enhanced vertical resolution. The newly revised BUFKIT was released to the forecasters in time for the record breaking 10 December 1995 LES event.

Figure 2 is an example of the opening BUFKIT screen which displays parameters for all model output levels for a given hour. The primary presentation, located on the right left side of the display, is the vertical sounding depicting temperature, moisture, and winds for the specific forecast hour. Inversions are shaded with the base height labeled. This display provides forecasters with an hour-by-hour presentation of key LES parameters including instability, fetch over the lakes, shear profiles, and height of the capping inversion. The 49-frame animation of this profile provides the forecaster with additional insight as the atmosphere evolves toward conditions supporting LES events.

The Lake Index (similar in concept to the Lifted Index) is an indication of the absolute stability of the lower atmosphere. BUFKIT displays to what extent the lower atmosphere is destabilized by the lake based upon forecast lake-air temperature differences. When this index becomes negative, the atmosphere has become absolutely unstable and is used by the forecast staff as an indication of the onset of LES band formation. The value of -5o shown in Figure 2 indicates conditions will become unstable by hour 28. Forecasters determine how high this instability extends by determining the height of the subsidence capping inversion. One of the conclusions from the LOWS project what the height of this capping inversion determines the intensity of the LES band (Reinking et al 1993). A capping inversion below 850 mb tends to minimize LES activity. The example in Figure 2 shows no capping inversion present below 500 mb at hour 28. Given these conditions, the forecaster could determine the potential for a very significant LES as high.

The left side of the BUFKIT displays an analysis of the low-level wind field to determine the location of LES band formation and migration. During the 1960s and 1970s the Buffalo forecast office prepared LES "threat regions" for a given wind direction over Lakes Erie and Ontario (Kolker 1978). These LES threat maps, available for every 10 to 20 degrees of wind direction, were valuable in identifying roughly what areas downwind of the lakes were at risk. BUFKIT extends this approach by "morphing" these threat areas for a specific model output wind direction. The example in Figure 2 shows BUFKIT placed the LES threat area in northern Erie County including the northern half of the Buffalo metropolitan area, for a wind direction of 252o.


The hourly animation capabilities of the profile display are effective in viewing the detailed evolution of LES bands. However, user feedback quickly identified the need to provide a time-height overview that summarizes all hours of the profile data. Figure 3 is an example of the BUFKIT overview display mode. This work area allows the forecaster to plot 22 LES parameters on as single display. The integration of these fields along a time-line provides the forecaster with additional insight into the critical LES parameters for this event.

The example in Figure 3 shows how forecasters can use the BUFKIT overview display to predict the dissipation phase of a LES event. Inversions are depicted as crosshatched boxes. The overview display shows the capping inversion rapidly descending after 12Z Monday. In addition, winds near 850 mb have veered from the southwest to the northwest seven hours earlier, effectively moving the LES bands south of Buffalo. Hourly animations of threat maps and helicity are still available on this display when the forecaster clicks and drags the mouse pointer across the time scale. Forecasters stated, it is this integration of these parameters that provided the needed insight to the onset, duration, and intensity of the LES bands.


At the conclusion of the 1995-1996 LES season, forecasters at NWSFO Buffalo and other NWS offices participating in the LES Project were asked to evaluate the effectiveness and LES tools and data sets. Figure 4 depicts the mean score NWSFO Buffalo staff gave LES tools used during the LES season. The forecasters found the BUFKIT display of early eta (48 km) forecast profiles was the most effective tool in forecasting LES. They stated that this high score could be attributed to two factors:

- the early eta profiles provided the vertical resolution needed to determine cap height, lake-induced absolute instability, and shear profiles below 700 mb.

- the data was available in a timely manner. Often, the data arrived early enough to allow forecasters to prepare a morning forecast instead of just a forecast update.

Profile data sets that failed to achieve this balance of resolution and timeliness were scored lower. The meso-eta version of BUFKIT was scored lower primarily because it was not routinely available until after the afternoon forecast were disseminated to the public and media.

The effectiveness of BUFKIT for LES forecasting were reflected in other Eastern Region forecast offices participating in the LES project. Forecasters in Eastern Region considered 1-hour model soundings to be the most critical data source for forecasting LES with the BUFKIT display application enabling forecasters to efficiently display and analyze this model output (Carter et al 1996).


The use of BUFKIT to display high resolution profile model output has significantly enhanced the forecasters' ability to predict the onset, duration, and intensity of LES bands. This application combines animation of critical LES parameters with ease of use to provide the forecaster with insight to the lake effect snow process. The Rapid Application Development language used by BUFKIT allows the program to continue change as model output capabilities and LES forecasting techniques evolve.


Carter, G. M. , T. L. Salem, and M. A. Wool, 1996: The use of new technology for enhanced snow forecasts in the vicinity of Lakes Erie and Ontario. Preprints, 15th Conf. on Wea. Anal. and Forecasting, Amer. Meteor. Soc., Norfolk, VA. 567-572.

Jensenius, J. 1990: Personal communication. NWS Techniques Development Lab. Washington, D.C.

Kolker, B., 1978: Current forecast procedures for lake effect snow in western New York especially related to the 1976-1977 and 1977-1978 winters. Proc. Of the 35th Ann. Eastern Snow Conf., Hanover, NH, 17-35. [Available from United States Army, CRREL, 72 Lyme Rd., Hanover, NH 03755]

Niziol, T. A. and J. M. Levan, 1992: Forecast procedures for mesoscale lake induced snow events in the eastern U.S., Preprints, Symposium on Wea. Forecasting, Amer. Meteor. Soc., Atlanta, 202-209.

_____, and E. A. Mahoney 1997: The use of high resolution hourly soundings for the prediction of lake effect snow. Preprints, 13th Intl. Conf. On Interactive Info. and Processing Sys. (IIPS) for Meteorology, Oceanography, and Hydrology, Amer. Meteor. Soc., Long Beach, CA. (elsewhere in this volume).

Reinking, R. F., R. Caiazza, R. A. Kropfli, B. Orr, B. E. Martner, T. A. Niziol, G. P. Byrd, R. S. Penc, R. J. Zamora, J. B. Snider, R. J. Ballentine, A. J. Stamm, C. D. Bedford, P. Joe, A. J. Koscielny, 1993: Lake Ontario winter storms (LOWS) project. Bull. Amer. Meteor. Soc., 74, 1828-1849.

Waldstreicher, J. S., T. A. Niziol, E. A. Mahoney, 1996: Lake effect snow forecasting - 1985 to 1995: A decade of scientific and public service advancements. Preprints, 15th Conf. On Wea. Analysis and Forecasting, Amer. Meteor. Soc., Norfolk VA.

Back to NWS Buffalo Research Index | Back to NWS Buffalo Home Page