1. Improving the Quality, Accessibility, and Relevance of Weather and Climate Data for Crop Insurance Christopher Daly Professor & Director, PRISM Climate Group NACSE/College of Engineering Oregon State University, Corvallis, OR

2. Topics Background and Introduction Brief overview of the PRISM Modeling System PRISM/RMA Web Portal

3. Northwest Alliance for Computational Science & Engineering Focuses on transitioning research advances into daily practice – Interdisciplinary research center founded in 1995 – On-campus collaborators from 6 colleges and 22 departments – Hosts web-based data/tools/portals for 30+ partnerships & consortia – All projects carried out collaboratively with industry and/or international, national, state, or not-for-profit organizations Entirely funded through “soft $$” (contracts, grants, etc.) Two focal areas for research – Climate mapping – Usability engineering

4. PRISM Climate Group Overview Applied research team at OSU PRISM climate mapping technology has been continuously developed, and repeatedly peer-reviewed, since 1991. Seminal paper (1994) cited in over 1,200 journal articles The PRISM Climate Group is the de facto climate mapping center for the US, and is the most advanced in the world Wide spectrum of sponsors reflects range of applications: USDA, DoC, DoD, DoE, DoI, EPA, NSF, ….

5. NACSE Usability Group Overview Applied research team devoted to usability engineering – Leader (Pancake) conducted seminal research on how scientists think about, and interact with, computers Institutionalized joint development of software tools by industry, research, and customer communities – Pioneered use of ethnographic techniques to identify user needs and preferences – Established world’s first lab for systematically testing software behavior across full spectrum of user desktop systems – Clients/partners have included computing industry (Intel, HP, IBM, Cray, Fujitsu) as well as agencies (DoD, DoE, DoEd, NSF, NOAA, ODOT, ODAS, USFS, USGS, TNC, etc.)

6. What Is Climate Mapping? The process of interpolating climate statistics at irregularly-spaced station locations to a regular grid “Geospatial Climatology” The study of the spatial patterns of climate on the earth’s surface and their causes

7. Why Are the Spatial Aspects of Climate Important? Weather and climate determine in large part what can grow where, and are the single most important causes of crop loss The location of interest is often not represented by a weather station Modeling, analyses and decision support tools are becoming increasingly spatial (GIS)

8. Why Is Climate Important? Climate provides a long-term context for weather events Weather is a variation on typical climate conditions The spatial patterns of long-term climate inform the spatial patterns of weather (“Climate Fingerprint”) 1971-2000 July Tmax 2003 July Tmax (One of Hottest on Record) Different values, but same spatial pattern

9. Weather Events & Climatological Context SingleYear Recent Past (< 30 yrs) Official “Normal” (30 yrs, 1971-2000) Climate Weather DailyMonthly Annua Annual

10. PRISM Parameter-elevation Regressions on Independent Slopes Model Generates gridded estimates of climatic parameters (e.g., P, T, DP) Moving-window regression of climate/weather values vs. elevation or climatology/radar for each grid cell – Uses nearby station observations Spatial climate knowledge base weights stations in the regression function by their physiographic similarity to the target grid cell

11. PRISM PRISM spatial climate knowledge base accounts for spatial variations in climate due to: Elevation – lapse rates Terrain orientation – rain shadows Terrain profile – terrain enhancement of precipitation Moisture regime – exposure to moisture sources Coastal proximity – marine air intrusion Two-layer atmosphere – inversion layer, free atmosphere Topographic position – susceptibility to cold air pooling

12. Station Networks Used in the PRISM Spatial Climate Datasets

13. PRISM/RMA Weather & Climate Portal New portal being developed for RMA – Provides fast access to quality climate data – Incorporates usability engineering so it’s easier to use and produces info relevant to RMA needs Initial prototype focuses on 4 key ways users need to find and apply climate data

14. Task 1: Check Recent Conditions View national-level precipitation and temperature patterns

15. Recent Conditions (2) View national-level precipitation and temperature patterns See how recent conditions compare with historical patterns

16. Task 2: Quick Summary for One Time/Place Select a particular location and time period – Entire continental US – Prototype: any 1-16 months since Jan 2006 Likely completeness of data for that date is indicated – Coming in future versions Earlier dates Daily data Longer time spans

17. Summary Assessment (2) Choose location – From county/PLSS list – By clicking on map – By typing lat/lon

18. Summary Assessment Results Select a particular location and time period Compare average conditions with 10- or 30-year data

19. Summary Assessment Results (2) Example: assessing conditions near Raleigh NC during 2010 tobacco season – Comparison with 30- year figures show unusually warm, dry conditions – Temps in 2010 actually exceeded any monthly averages recorded 1971-2000

20. Summary Assessment Results (3) “Typical,” “wet,” “unusually wet,” etc. provide qualitative assessments based on algorithmic rank analysis

21. Summary Assessment Details Details on each year’s conditions and ranking are available as an option

22. Task 3: Explore Detailed Data View plots of time- series data for a selected location Example: block in southwestern Fresno County, CA Period starting October 2010

23. Detailed Data (2) Interactive: mouseover reveals specific values user selects which lines are visible

24. Detailed Data (3) Temperature values & 30-year comparison also available Data values downloadable as spreadsheet file

25. Task 4: Generate Customized Report Select a 16-month "insurance period" and location Get an on-demand prevented planting Future plans – Other types of reports – Other systems of legal identifiers

26. Customized Report (2) Example: – Selected location in Brookings County SD by clicking on map – Need prevented planting report for period leading up to May 2010

27. Customized Report (3) Report generated “in real time” Compute-intensive – Searches and analyzes 1000s of grids, each containing 1000s of data values – Requires high- performance data storage and computers

28. Customized Report Results Report style based on manually-created versions that were used successfully in court cases Generating usable English-language text was a huge challenge

29. Customized Report Results (2) Features dynamically created maps and tables as well as text Interactive tables allow user to sort based on, e.g. – Distance from parcel – # observations

30. Customized Report Results (3) Interactive tables facilitate scrutiny

31. Customized Report Results (4) Science-based analysis uses internationally respected PRISM climate model Chosen by USDA for new updates to Plant Hardiness Zones

32. Science Based, but Designed for Usability Features to maximize user efficiency Examples Sensible “default values” – User may not even have to choose a value No typing for normal tasks – Only needed for customizing titles or specifying lat/lon Minimal steps needed to accomplish any task Features to minimize user errors Examples Explicit information about potential mis-use – Data quality & completeness – Methods used Little or no typing – Almost eliminates possibility of typos Descriptive text – not just data