Understanding the needs of decision makers: an approach to develop better products and Services

Transcription

1 Understanding the needs of decision makers: an approach to develop better products and Services By Serge Desjardins* Bill Appleby** WWOSC 2014 Montreal, Canada 18 August 2014 * National Lab for Marine and Coastal Meteorology, Meteorological Service of Canada, Halifax, Canada ** Products & Services Development, Meteorological Service of Canada, Halifax, Canada

2 PLAN - Information versus knowledge - Evolution of the data (its accessibility & interpretation) - Services from Science (Packaging the information) - Client Oriented & Collaborative Culture (examples) - Conclusions

3 Information vs knowledge Age of information does it mean age of knowledge? Are we good swimmers?

4 Information vs knowledge Age of information does it mean age of knowledge? Are you a good swimmer? 6 feet deep YES NO No problem I will give you access to a better pool.

5 Information vs knowledge Age of information does it mean age of knowledge? Now, Are you a good swimmer? 3 feet deep 8 feet deep YES NO, I am afraid not You might need some lessons Or a life jacket then

6 Information vs knowledge Age of information does it mean age of knowledge? SWIMMING POOL ANALOGY Good Swimmer - lessons : education Pool (water) : INFORMATION - training : forecast learning Various Depths : - practice : forecasting - Formats (media, web, maps, text) - Complexity (deterministic, probabilistic) - resolution (temporal, spatial) SWIMMER : Use of the Information : KNOWLEDGE WHY A BAD SWIMMER STAYS A BAD SWIMMER? A user not using properly the information Knowledge comes from education and experience Information can be useless or even misleading if not used properly

7 Information is a good thing when used properly Various depths in the pool : - versatility of swimmers (skill & knowledge in public) - versatility of products and services (from simple to more complex) PRODUCTS more details in high impact weather COMPLEXITY 8 feet 3 feet information General Weather Patterns 8 feet deep USERS Decision makers Client needs 3 feet deep Public

8 Information is a good thing Why we want to swim more in the deep area of the pool observations Hourly & gridded site point forecast Longer term forecast Probabilistic forecast Interactive Maps 2D : single to multiple fields 2D: probabilistic envelops 1D : Time series 1D: Meteograms - Higher model resolutions - More specialised forecasts Need to supply MORE LIFE JACKETS because not all of our clients will ever be good swimmers (not enough knowledgeable to fully use the information)

9 Information is a good thing with knowledge Need to package the information for the users (based on their needs) Type of users (Decision) : Public A: Make the decision for me : Supplying a scenario (public in general) B: I want to know the alternatives : scenario with probability C: Just give me the information : enough knowledge to build his/her scenario Type of users (Decision makers) : understand the making decision process I: Impact : How important/critical Weather is for the client II: Threshold dependencies : The values that trigger a decision III: Accuracy : sensitivity to inaccuracy

10 SCIENCE TO SERVICE SERVICE FROM SCIENCE Weather Related Across the board -Winter Storm - Summer Storm -Arctic - Air quality -Marine - Coastal flooding - Hurricane - River Flood - Avalanche - Forest Fire - Transport - etc SERVICE FROM SCIENCE

11 Evolution of the Data and its accessibility and its interpretation FORECAST BULLETINS & MODEL OUTPUTS Information Before 2010 Now & Future? PUBLIC Knowldege (interpretator) Packaging CLIENTS NEEDS Decision Makers Gov. Private

12 SERVICE FROM SCIENCE Numerical Guidance Research Meteorological Service of Canada Information APPLIED SCIENCE BRIDGE-UNITS SERVICE-ORIENTED Operational unit Forecast & Services Forecasters Warnings Public Format WPMs Innovative unit Products & Services Development Specialized format CLIENTS Knowldege (interpretator) Packaging

13 CLIENT ORIENTED & Collaborative culture FREEZING SPRAY FORECAST «We think we know what they want or what they need» Development of more sophisticated freezing spray algorithms «When freezing spray is mentionned in the forecast we do not take any chance and stay home» from fishermen with small vessels (ex: lobster boat type) OUR ROLE in that case Are we overforecasting..? Research - Identify the challenge for us (vessel size, speed) : which client? - might need more sophisticated freezing spray algorithms - might need research Service - need observations from the client. client

14 CLIENT ORIENTED & Collaborative culture STORM SURGE Does giving a total sea level value satisfy the client? «We need to know the ptoential flooded area and the critical period when it could happen, knowing the uncertainty from model. OUR ROLE in that case Phase 1 Make sure we have a good & efficient storm surge model Develop a coupled tidal-storm surge-wave model system : real total sea level Determine the treshold values and their sensitivity (impact) from client Test the products with clients (EMO & municipality) before implementation. Research Service Phase 2 Develop a probabilistic approach in forecasting coastal inundation. Clients Test with the clients (EMO & municipalities) Indentation maps

15 CONCLUSIONS The increase of information can be seen as a challenge but also an opportunity to give better weather service to the public Careful about the access of the information : more info more knowledge Because there is no real increase of knowledge (in the increase of information) more information means more interpretation, more packaging To better package : BE client oriented must DEVELOP a collaborative culture (decision makers) must UNDERSTAND the decision making process from the client A bridge must exist between Science & Service interpretation and packaging of the information for the user in MSC: Applied science bridge-unit : Products & Services Development

16

17 CLIENT ORIENTED & Collaborative culture SEA STATE FORECAST We can give a better forecast by being more accurate in the wave height and the period and by supplying swell components and wave spectrum information «Winds do not stop us neither waves. What we need to know is how the boat will be shaken for planning and functional purposes. OUR ROLE in that case Phase 1 Make sure we have a good wind field that drives a good & efficient wave model. Develop a dangerous sea index product (wave height and period, mixed sea) Test the product with clients (marine community) before implementation. Phase 2 Develop a probabilistic modeling system for wave Develop probabilistic products Test with the clients (marine community) Service Research Clients Dangerous Sea maps (steepness, mixed sea)