Survey: Why Limited Use of Ensembles in Operations?. Stony Brook University-SUNY CSTAR Dr. Brian A. Colle [email protected] Purpose: To rank the potential issues of using ensembles in operations within the National Weather Service. Survey Questions:
Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.
Stony Brook University-SUNY CSTAR
Dr. Brian A. Colle
What is your current position? (operational forecaster, manager/admin, researcher, and model developer)
How often do you use ensembles?
Why are ensembles not used in operations as much as they could be? Weight each issue from “not a problem” (= 1) to “one of the largest problems” (= 5)
Rank the 8 provided issues from the largest problem to the smallest issue/problem.
Open Ended Question: Other written comments regarding the difficulty in using ensembles in operations.
* The results in the following slides are separated into the following sections. Number of responses are in the ( )
All responses (166)
Operational Forecasters (111)
Admin / Managers (37)
Researchers and Model Developers (18)
Forecasters who sometimes or never use ensembles (50)
Forecasters who use ensembles often or always (61)
* For each section above, there are two sets of plots for questions # 3 and 4 to show the average value/rank and another to show the distribution of responses.
2 = 35%
1 = 31%
2 = 36%
Why Ensembles not used in operations as much as they could be?
(1=Not a problem, 2 = Minor, 3= Moderate, 4 = Major, 5= One of Biggest)
Top Median Response Number Also Shown and Percent
2 = 30%
3 = 32%
3 = 31%
3 = 32%
#8 = 19%
#8 = 25%
#5 = 19%
Average Rank of the issues (from 1 to 8) limiting ensemble usage in operations
Top Median Rank Number Also Shown and Percent
#4 = 19%
#5 = 17%
#6 = 17%
#2 = 32%
Nearly all forecasters use ensembles; however, only ~45% use them “sometimes.”
Each ensemble issue has a large distribution of rankings (some problems were considered major for some and minor for others) – so there are a variety of opinions as illustrated by the open ended questions.
The highest ranked issues are (1) the lack of tool/graphics, (2) ensemble data access, and (3) ensemble training.
For forecasters who “sometimes use ensembles,” it appears that the time to view/interpret ensembles is also an issue (ranked #3). Some of this may be from the lack of good tools and data access as suggested by their open ended questions.
Managers/admin view the issue of time to view/interpret ensembles as the smallest problem on average (ranked #8).
Most or all researchers and model developers use ensembles. They clearly believe that the lack of tool/graphics is the largest issue.