Providing Weather Forecast Uncertainty Information to the Public

1
Providing Weather Forecast Uncertainty
Information to the Public

• Julie Demuth, Rebecca E. Morss, Jeffrey K. Lazo
• NCAR Societal Impacts Program
• Boulder, CO

2008 WASIS Workshop August 12, 2008
USWRP
2
Suppose the forecast says, There is a 60 chance
of rain tomorrow.
What do you think best describes what the
forecast means?
Percent of respondents
 Response option
It will rain tomorrow in 60 of the region.
It will rain tomorrow for 60 of the time.
It will rain on 60 of the days like tomorrow.
60 of weather forecasters believe that it will
rain tomorrow.
I dont know.
Other (please explain)
3
Uncertainty research questions

• How much confidence do people have in different
  types of weather forecasts?
• Do people infer uncertainty into deterministic
  forecasts and, if so, how much?

• How do people interpret a type of uncertainty
  forecast that is already commonly available and
  familiar probability of precipitation forecasts?
• To what extent do people prefer to receive
  deterministic vs. uncertainty-explicit forecasts?
• In what formats do people prefer to receive
  forecast uncertainty information?

4

• How do people interpret a type of uncertainty
  forecast that is already commonly available and
  familiar -- probability of precipitation (PoP)
  forecasts?

• 90 of respondents received close-ended version
  of the question
• 10 of respondents received open-ended version

5
Suppose the forecast says, There is a 60 chance
of rain tomorrow.
What do you think best describes what the
forecast means?
Percent of respondents
 Response option (N1330)
It will rain tomorrow in 60 of the region.
It will rain tomorrow for 60 of the time.
It will rain on 60 of the days like tomorrow.
60 of weather forecasters believe that it will
rain tomorrow.
I dont know.
Other (please explain)
Technically correct interpretation, according
to how PoP forecasts are verified (Gigerenzer et
al. 2005)
6
Open-ended responses re PoP

• Interesting insight from the responses to
  open-ended question and other write-in
  responses
• Many reiterate PoP without clarification
• Many describe the chance theyll personally
  experience rain or personal implications for
  action
• Consistent with other studies, majority of people
  dont know technically correct definition of PoP
• but asking people to think about PoP from a
  meteorological perspective may have limited value
  people still have to infer what it means to
  them!

7

• To what extent do people prefer to receive
  deterministic vs. uncertainty-explicit forecasts?
• In what formats do people prefer to receive
  forecast uncertainty information?

8
All the choices below are the same as a
probability of precipitation of 20.
Do you like the information given this way?

• Chance of precipitation is 20
• There is a 1 in 5 chance of precipitation
• The odds are 1 to 4 that it will rain
• There is a slight chance of rain tomorrow

? Percent ? Frequency ? Odds ? Text
Asked this question 3 ways -- using PoPs of 20,
50, and 80 with corresponding text descriptions
from NWS
9
Percent of respondents who said yes
10
Suppose the high temperature tomorrow will
probably be 85ºF. A cold front may move through,
making the high only 70ºF.

• Would you like the forecast given this way?

The high temperature tomorrow
will be 85F
11
Percent of respondents who said yes"
Deterministic 35
Deterministic only 7
Uncertainty 90
Uncertainty only 63
N1465
12
Discussion Topics

• Balance between educating the users and
  understanding users needs, wants, uses?
• Proliferation of weather information, different
  sources, different media, etc.
• Peoples desire for getting consistent info
• On-air constraints for broadcasters