Close Window

Desert Research Inst.: "Improved snowfall QPF in radar-limited terrain"

Final Report


Field research and model forecast verification related to quantitative snowfall forecasting was conducted in the Park Range of the Colorado Rocky Mountains during two winter periods (January-March 2002 and December 2002). The partner institutions were the University of Nevada Desert Research Institute (DRI) and the National Weather Service Forecast Office in Grand Junction, Colorado (GJT). The field research and analysis included measurements of cloud and precipitation microphysical parameters, vertical profiles of thermodynamic and air motions, snowfall density and precipitation rates, cloud and precipitation chemistry, meteorological conditions at multiple elevations, and numerical results from a mesoscale prediction model.

The field program was based at the DRI Storm Peak Laboratory (SPL), a high-elevation (3,200 m ; 10,500 ft MSL) research facility in the Park Range of northwest Colorado which provides an ideal site for cloud physics and snowfall research. A concurrent NSF-sponsored research program being conducted by DRI provided additional data from a vertically-pointing radar, frequent balloon soundings and other observations from a valley elevation on the upslope (western) side of the Park Range near Steamboat Springs, Colorado.

Primary accomplishments of this project include development of procedures for more effective snowfall forecasting in mountainous terrain when radar coverage is not available, quantitative analysis of the accuracy of snowfall predictions from mesoscale model forecast products at multiple elevations, characterization of orographic snow event types that commonly occur in the study region, and improvement in observational data available for assessing winter weather conditions in the Park Range (Wetzel et al., 2002; Wetzel et al., 2003).


The project involved the teamwork of three scientists from GJT (SOO Dr. Mike Meyers, Paul Frisbie and David Nadler) with three scientists from DRI (faculty Dr. Melanie Wetzel and Dr. Randolph Borys, and research assistant Andrew Rossi). In addition, there was excellent collaboration with faculty, scientists and students from Colorado State University. Four meetings were held in conjunction with the Partners Project, at GJT and SPL.

In cooperation with Dr. William Cotton from Colorado State University (CSU) and Dr. Meyers from GJT through another COMET Outreach Project, a high-resolution numerical prediction model was used to investigate the cloud and microphysical conditions during the field project. The CSU RAMS (Regional Atmospheric Modeling System) provided mesoscale nested forecast grids, soundings and time series of snow, cloud, wind and thermodynamic parameters. The RAMS finest grid was centered over SPL on the Park Range. In addition to direct forecast and research support for the mountain-weather field project, the RAMS forecasts were beneficial to the day-to-day forecast operations at GJT. During the field project, the NCAR MM5 model was employed. This archived data is available for a more thorough post analysis investigation.

Additional collaboration with SPL not directly related to the COMET project included two separate high-wind events that occurred near Steamboat Springs (Meyers et al, 2003; Jones et al. 2003). Observational data was provided to GJT for these events by the scientists at DRI / SPL. The project also contributed to a separate DRI research program sponsored by NSF to investigate aerosol effects on snow crystal growth. This program was conducted at SPL by DRI during the same time period as the Partners Project, and datasets collected by each project were utilized to expand the accomplishments of both.


Publications and conference presentations that were directly produced as a result of this project, or which contain results that benefited from this Outreach partnership, are listed below:

Borys, R.D., D.H. Lowenthal, S. Cohn, W. Brown, 2002: Mountain-top and radar measurements of the effects of aerosol on snowfall rate. AMS 12th Conf. on Mountain Meteorology, Park City, UT.

Jones, C.N., J.D. Colton, R. McAnelly, and M.P. Meyers, 2003: A mountain wave event west of the Colorado Park Range. Nat. Wea. Dig. In press.
Meyers, M.P., J.S. Snook, D.A. Wesley, G.S. Poulos, 2003: A Rocky Mountain storm - Part II: The Forest Blowdown over the West Slope of the northern Colorado mountains-observations, analysis and modeling. Wea. Forecasting. In press.

Wetzel, M., M. Meyers, R. Borys, A. Rossi, R. McAnelly, W. Cotton, P. Frisbie, D. Nadler, S. Cohn, W. Brown, 2002: Verification of snowfall forecasts in the Park Range of Colorado. Ninth Intermountain Weather Prediction Workshop, Salt Lake City, UT.

Wetzel, M., M. Meyers, R. Borys, R. McAnelly, W. Cotton, A. Rossi, P. Frisbie, D. Nadler, D. Lowenthal, W. Brown and S. Cohn, 2003: Mesoscale Snowfall Prediction and Verification in Mountainous Terrain. In preparation for Wea. Forecasting (expected submission date 15 April 2003).


4.1 A primary benefit of this project for the university research group is an improved knowledge of synoptic and mesoscale atmospheric conditions that lead to heavy snowfall events at SPL and in the Park Range in general. It has been extremely valuable to be able to discuss and evaluate specific snowfall events with operational forecasters, in order to enhance the goals and success of the field research projects frequently conducted at SPL. The results of the forecast model validation efforts, increased familiarity with NWS operational prediction methods, and new knowledge on snow growth processes and orographic effects collectively have enhanced the research capabilities and opportunities for the participating faculty.

Through the initiatives of the research, we have also been able to augment a graduate-level course (ATMS 792) from University of Nevada, Reno (UNR). The subject of this course is field training in atmospheric instrumentation and research methods for mountain meteorology, and it is offered at SPL by the UNR/DRI faculty during the winter storm season. The course that was conducted during January 2002 with graduate students in degree fields of Atmospheric Sciences, Hydrology, Chemistry, and Environmental Sciences from UNR and Univ. of Nevada, Las Vegas (UNLV). A major focus of the course is development of an independent field research project by each graduate student that integrates aspects of winter weather forecasting with applied research topics (such as snow hydrology, cloud microphysics and air pollution-cloud interactions). The addition of RAMS model products tailored to the SPL location and the Park Range high-resolution grid was a very valuable addition to the UNR/DRI research and teaching activities, and for other college classes taught at SPL (undergraduate classes from Wisconsin and New York).

NWS forecasters involved in this COMET Partners Project took part in the field measurement period at SPL with the DRI scientists, gaining first-hand knowledge of the profiling radar, cloud microphysical sampling, snow survey, and other aspects of the field research campaign. The partnership between DRI scientists at SPL and the NWS team at GJT has improved the training of forecasters as well as students, and has also led to the development of new collaborative research initiatives.

4.2 Operational forecasting for the Park Range of north central Colorado is problematic due to WSR-88D distance and beam blockage issues, and the sparsity of surface observations in the region. This region of GJT's county warning area (CWA) often has shallow but fairly intense winter storms where extreme precipitation variations occur over relatively short distance as a result of the complex topography of the region. This project has put this region under a microscope, with more detailed observations and insights from the DRI scientists, as well as high-resolution cloud modeling from the CSU RAMS. Collecting data from two consecutive seasons has allowed the forecasters to examine a wide spectrum of case studies, which have been investigated in both real time and in research mode after the events.

One specific forecast problem that often occurs in Park Range is a shallow upslope (northwest flow) where seemingly low specific humidity is available. In one such case, 15-16 January 2002, the GJT forecasters expected 3-6 inches snowfall for the mountain locations which is just below snow advisory criteria. Detailed observations from the project showed a larger gradient of snowfall based on elevation with 1-3 inches across the Steamboat Springs town to over one foot at the top of Steamboat ski area. In this event the GJT forecasters did quite well forecasting the QPF, however, actual snowfall forecasts were inaccurate due to the extremely low snow density in this event (3% density observed). As a result of this storm and similar subsequent events during this project, the GJT forecasters are more cognizant of important microphysical and topographical effects which can impact snow density and ultimately total snowfall.

This project has provided the research framework for GJT forecasters to examine and compare the high-resolution forecasts from the RAMS model to other operational models developed at NCEP. The analysis and evaluation have shown the importance of the model physics incorporated to RAMS, in addition to finer grid spacing, to obtaining value-added forecasts over complex topography. The RAMS model with its more precise microphysical scheme has proven to be a very useful research tool in post-event comparisons with the detailed observational measurements provided by the DRI scientists.

The COMET Partners Project collaboration has facilitated the incorporation of the WES into the forecast environment. GJT forecasters have archived numerous cases for these COMET applications that facilitate research and training, and the system has proven to be an invaluable tool for post-analysis of specific storms. Data from the WES archive, with additional field measurements collected by DRI and NCAR, and the RAMS gridded model output, were critical in carrying out this research project and supported the collaborative authoring of a journal article soon to be submitted for publication (Wetzel et al., 2003; draft manuscript attached to this report).