Close Window

Pennsylvania State University: "Operational ensemble forecasting and data display techniques"

Final Report


The primary objectives of this study were to provide ensemble products specially tailored for WFO operations and assist operational forecasters transition from deterministic forecasting to probabilistic forecasting. An important aspect of this study was to demonstrate the advantages in skill of ensembles relative to any single forecast.

During the twelve months of this project, we successfully integrated web-based ensemble forecast products into operations at the NWSFO in State College. Additionally, through email and two visits to surrounding offices, we were able exploit these technologies and concepts. These data are routinely used in the preparation of both long and short-term forecasts in many Eastern Region and several Central Region Offices. These data have been of great value in the preparation of the 6 and 7-day forecasts.

The use of historical climatological data with the ensemble forecast data has provided a very useful forecast tool A comparison of the ensemble consensus forecasts to the 30-year Climatology indicates that the consensus shows skill in forecasting significant weather events. It appears that strong departures from normal at longer ranges (beyond about 6 days) provide a measure of predictability and increase the forecaster's confidence in the forecasts. This work has been documented in two journal papers (accepted for publication) and in several conference papers.

Applying these 30-year climatological fields to the individual forecast members has revealed a means to identify significant weather events at shorter ranges. The NWS offices in State College, Boston, and Syracuse Indian, have been able to identify several significant weather events in advance. The spin-off of this, since mid-August, has been an expanded use of these data by many Eastern Region forecast offices. These anomalies in the ensemble have recently appeared in forecast discussions from the NCEP extended range-forecast group. During the record East Coast trough of May 2001, HPC spoke about the intensity of the anomalies and the number of standard deviation the trough departed from normal.

Significant project accomplishments included: the construction of real-time WebPages with displays of ensembles and the departures from Climatology, presentations at the October 2000 NWA; two posters presented at the American Meteorological Societies Quantitative Precipitation Symposium (Jan 2001), and two papers, accepted for publication in Weather and Forecasting and Monthly Weather Review.

Additional accomplishments include:

a. An archive of 50-years of climatological data for future research on significant weather events.
b. An on-line 35GB archive of 12 months of NCEP ensembles.
c. WebPages showing case studies of the use of these data.
d. A presentation at the Pittsburgh, PA WFO winter weather workshop. (Richard Grumm)
e. A presentation at the Binghamton, NY WFO winter weather workshop. (Robert Hart)
f. A paper, under construction on the use of ensembles.
g. Significant snowstorm case studies using these data at the Syracuse Indiana WFO (Julie Adolphson)
h. The inclusion of these data in a course on long-range forecasting during the spring semester at the PSU. The course was co-instructed with Mr. Paul Knight (PSU) and Richard Grumm(NWS). Several WFO and MARFC employees attended the course.

SECTION 2 Summary of Exchanges

2.1. University

Robert Hart presented the forecast method, data, preliminary results, and examples to the National Weather Service Office in Binghamton, New York on 30 November 2000. This led to the sharing of knowledge between PSU and the NWS.

Robert Hart also has authored and co-authored several papers related to the project. The paper, entitled Using Normalized Climatological Anomalies to Rank Synoptic-Scale Event Objectively by Robert E. Hart and Richard H. Grumm will appear in MWR this autumn. Robert Hart took the initiative to complete this paper and ensure its publications.

2.2 NWS Partner

The NWS implemented ensemble training in the office. This training included sessions related to ensembles and consensus forecasting at the Winter Weather Workshop and operationally oriented use of the ensemble graphics in the forecast process. This has raised the office's overall awareness about ensembles and the value of the consensus forecasts when compared to the 30-year Climatology. PSU staff members and students were invited to these presentations and subsequent presentations wherein it was demonstrated how to use the ensemble and climatological data to anticipate major convective weather events.

In addition to the local gains here in State College, the Boston NWS office has also made extensive use of these data. Walt Drag (Lead Forecaster) presented case examples of these data at the Canadian Maritimes Weather Centre in Dartmouth, Nova Scotia, October 2000. A dialog on the use of these data between KBOS, Richard Grumm, and Robert Hart evolved out of this.

The paper, entitled: Standardized Anomalies applied to significant cold season weather events: Preliminary Findings by Richard H. Grumm and Robert E. Hart was accepted for publication in Weather and Forecasting. This paper should appear in WAF during the winter of 2001-2002.

SECTION 3 Presentations and Publications

3.1 Presentations:

The preliminary results and case examples have been presented at the following venues:

a. COMET QPF Symposium September 2000
b. National Weather Association Meeting 9-13 Oct 2000
c. Pittsburgh National Weather Service Office Winter Weather Workshop 19 October 2000
d. State College National Weather Service Office Winter Weather Workshop 17 November 2000
e. Binghamton National Weather Service Office Winter Weather Workshop 30 November 2000
f. AMS QPF Symposium January 2001. Two papers and posters.
g. Spring 2001 Severe Weather Workshop, April 2001.

3.2 Publications:

Grumm, R.H. and R. Hart, 2001: Anticipating heavy rainfall: Forecast aspects. Preprints, Symposium on Precipitation Extremes: Prediction, Impacts, and Responses, Albuquerque, NM, Amer. Meteor. Soc., Boston, MA.

Grumm, R.H., and R. Hart, 2001: Standardized anomalies applied to significant cold season weather events: Preliminary findings. Accepted for publication in WAF to appear in a late 2001 issue

Hart, R., and R.H. Grumm, 2001: Anticipating heavy rainfall: Climatological aspects. Preprints, Symposium on Precipitation Extremes: Prediction, Impacts, and Responses, Albuquerque, NM, Amer. Meteor. Soc., Boston, MA.

Hart, R., and R.H. Grumm, 2001: Using normalized climatological anomalies to rank extreme synoptic-scale events objectively. To appear in MWR Autumn 2001.

Grumm, RH. and Robert Hart: The use of ensembles and 30-year climatological data to forecast significant weather events. In progress.


4.1 University

Students and staff of the Pennsylvania State University have gained knowledge on the use and display of ensemble forecast products. This information has been included into forecast discussions and was included in a course offered during the spring 2001 semester.

Other benefits include better interaction with NWS employees and enhanced information and graphics for use in University media outlets (such as its daily television show, Weather World).

A new course offered this coming autumn will use these data to forecast threats in the eastern United States in the 3-7 day forecast period. The threat forecasts will begin in July and the forecast class will begin in August of 2001.

4.2 NWS Partner

The NWS has benefited greatly from this project both locally and regionally. These data have helped forecasters better anticipate significant weather events and identify these events well in advance. Several recent success stories include:
· Anticipating the potential anomalous precipitation event of 12-13 August 2000
· Identifying the potential severe weather event over Ohio and Indiana in September 2000.
· Anticipating the early arctic surge and record cold of 7-9 October 2000.
· Anticipating the cold surge of Nov-Dec 2000
· Anticipating the major East Coast cyclone of 5-7 March 2001
· Anticipating the anomalous trough of May 2001.

The ensemble and ensemble departures from normal graphics, available at, are widely used by the National Weather Service Office in State College and other Eastern Region offices. Several Central Region offices, including Syracuse, Indiana, routinely access these data for daily forecast activities.

Collectively, the NWS has less then one year's experience using these data. The potential spin-off studies and applications using these data have yet to be fully realized.

In addition to the forecast applications now being employed, the NWS has gained experience in the use of GrADS software due to its association with PSU. Other benefits include access to model data via the PSU Internet connection and the availability of the forecast products on two webservers including: and .

Finally, Dr. J.M. Fritsch made a presentation on the value of model Consensus forecasts at the State College Winter Weather Workshop. This talk, combined with the ensemble-forecast tools produced from this project helped increase the awareness of the operational forecast staff on the value of ensembles and ensemble consensus forecasts. More progress in this area is expected this coming autumn.


No serious problems have been encountered with this project to date. However, we had to make two large disk purchases to ensure space for the exceptionally large databases. Also, we were unable to obtain SREF data during this project but continue to pursue this avenue. As of 21 June we had obtained limited access to the NCEP SREF data and have just begun using these data. We were unable to get other offices to participate in sharing mesoscale model data for ensembling.