The WxEM Project

Wave 6: Emergency Manager Use of Probabilistic Forecast Information

Motivation

  • NWS forecasters communicate with core partners throughout the forecast and warning process to ensure that the information they provide is relevant, understandable, and actionable. Emergency Managers are among the most important partners. They are a core linkage between NWS information and community actions that save lives and protect property.
  • Effective communication with EMs requires accurate and timely information about hazards (forecasts and warnings) and knowledge about types of information that EMs require when making decisions.
  • The WxEM project will provide this information by identifying and talking with a diverse group of EMs who are willing provide routine feedback on new technologies, forecast products, and NWS decision support services.
    • Surveys
    • Experiments
    • Interviews

Wave 6: Emergency Manager Use of Probabilistic Forecast Information

  • Objective: Understand how Emergency Managers use probabilistic forecast information
    • Do EMs use probabilistic information?
    • What are some probabilistic forecast preferences?
  • Do Emergency Managers use forecast-based decision thresholds?
    • What types of thresholds do they use?
    • Do they use probabilistic forecast information for these thresholds?
  • These slides provide basic information about the sample and highlight results from the Wave 6 survey
    • Over 230 responses

Survey Results

Geographic distribution of respondents

Survey Results

The NWS often uses probabilistic information (i.e., the likelihood of an event happening) to describe the possibility of a weather event occurring, particularly when there is still uncertainty in the forecast.

Do you, personally, use probabilistic information from the NWS?

Survey Results

The NWS often uses probabilistic information (i.e., the likelihood of an event happening) to describe the possibility of a weather event occurring, particularly when there is still uncertainty in the forecast.

How useful is probabilistic information from the NWS?

Survey Results

You indicated that you do not use probabilistic information from the NWS. Can you tell us why? Please select all that apply. [n=7]

Survey Results

Now we are going to ask you some questions about some NWS forecasts and graphics that contain probabilistic forecast information. This survey is about how probabilistic information is presented and how you as an emergency manager might interpret that information. There are no “right” or “wrong” answers, but rather we are looking at the clarity and usefulness of these forecasts and graphics. So, do not worry if you haven’t seen a graphic or forecast before, just give us your first impressions and interpretations.

Participants were shown one of these two graphics:

Survey Results

The graphic above represents uncertainty in the low temperature forecast. Based on your interpretation, which of the following choices best describes the forecast?

Survey Results

Based on the graphic above, what do you think is the most likely low temperature for Waco?

Survey Results

If you received this graphic from your local NWS office would you share it with key decision-makers and/or members of the public? Please select all that apply.

Survey Results

Now we have some questions about probabilistic snowfall forecasts.

Forecasts for snowfall can be provided as a range of potential values. Typically, forecasts with wider ranges of snowfall amounts are more likely to include the actual amount of snow that will fall. And vice versa, forecasts with narrower ranges of snowfall amounts are less likely to include the actual amount of snow that will fall.

Participants were shown one of the two following questions:

Survey Results

Imagine there’s a winter storm forecast for tomorrow and that you see all the snowfall forecasts below, which are different ways of presenting snowfall ranges and the likelihood that the actual amount will fall in that range. Which of these options do you prefer?

Survey Results

Imagine there’s a winter storm forecast for tomorrow and that you see all the snowfall forecasts below, which are different ways of presenting snowfall ranges and the likelihood that the actual amount will fall in that range. Which of these options do you prefer?

Survey Results

Imagine a situtation where you see a forecast but the forecast is incorrect as described below. HOw likely would you be to use forecasts from that source in the future?

Survey Results

In addition to text information, snowfall forecasts often include graphics indicating the amount of snow that is likely to fall in different areas, like the one below:

Participants were shown one of these two graphics:

Survey Results

How easy is it to understand the information presented in this graphic?

Survey Results

If you were an emergency manager for Springfield, how useful do you think this graphic would be for snowfall events?

Survey Results

What channels do you use to disseminate forecast information to your partners? Please indicate all that apply.

Survey Results

The NWS has also been testing another way to provide forecast snowfall amounts, similar to what is shown in the graphic below:

Survey Results

How easy is it to understand the information presented in this graphic?

Survey Results

If you were an emergency manager for Syracuse, how useful do you think this graphic would be for snowfall events?

Survey Results

If you received this graphic from your local NWS office would you share it with key decision-makers and/or members of the public? Please select all that apply

Survey Results

Imagine you see one of the following messages for your area from an NWS forecaster.

How easy is it to understand the information presented in this message?

Survey Results

How concerned would you be based on the information presented in this message?

Survey Results

If you received this message from your local NWS office would you share it (in this form) with key decision-makers and/or members of the public? Please select all that apply

Survey Results

You may have seen the following graphics from your local NWS office. They show two different ways to visualize probabilistic forecast information:

Survey Results

If you had to choose, which of these two probabilistic forecast formats do you prefer?

Survey Results

As you may know, the NWS Weather Prediction Center (WPC) issues Excessive Rainfall Outlooks that indicate flash flooding threats across the contiguous United States. These Outlooks use a combination of words, colors, and numeric probabilities to represent flooding threats. Emergency Managers were shown one of the following examples, one where the highest category was slight or one where the highest category was high:

Survey Results

Have you ever seen an Excessive Rainfall Outlook before?

Survey Results

How confident are you in your ability to understand the information presented in this Outlook?

Survey Results

Imagine you are an Emergency Manager for Location A… Emergency Managers were shown one of the following examples, one where the highest category was slight or one where the highest category was high:

Survey Results

Based on this Outlook, would you say that Location A is…

Survey Results

Based on this Outlook, how concerned are you about flash flooding at Location A?

Survey Results

Based on this Outlook, how likely are you to take any actions regarding potential flash flooding in Location A?

Survey Results

Based on this Outlook, how likely are you to take any actions regarding potential flash flooding in Location A?

Survey Results

To what extent does working with collaborators make you more capable of engaging in recovery activities following a hazardous weather event?

Survey Results

Now let’s look at some probabilistic wind forecasts. Imagine you are an emergency manager at location Y, and you are particularly concerned about the possibility of 74 mph winds. Your local NWS office provides this forecast graphic. EMs were shown one of these two graphics: ## Survey Results {.smaller}

How easy is it to understand the information presented in this graphic?

Survey Results

How concerned would you be based on the information presented in this graphic?

Survey Results

If you received this graphic from your local NWS office would you share it with key decision-makers and/or members of the public?

Survey Results

Some emergency managers have thresholds for decision-making. For example, opening warming shelters when the forecasted low temperature is 32° F or lower, or opening cooling shelters if the high temperature exceeds 100° F. Does your jurisdiction have specific weather thresholds for decision-making?

Survey Results

Have these decision threshold(s) been beneficial to your operations?

Survey Results

Is your local NWS office aware of your decision threshold(s)?

Survey Results

Did you work with your local NWS office to establish that threshold?

Survey Results

Does your local NWS office provide information tailored to that threshold to help in your decision-making process?

Survey Results

Does your local NWS office provide probabilistic forecast information tailored to that threshold to help in your decision-making process?

Survey Results

Does your emergency management office have any written plans or policies that include those thresholds?

Survey Results

Do you think establishing a decision threshold could be beneficial to your operations?