1. Methods to improve Real-Time Visualization and Exploration of Precipitation and Temperature in Web-Cartography ICC 2009, Santiago de Chile Christophe Lienert, ETH Zurich

2. Overview  Motivation  User needs and objectives  Methodology – automated workflows for P and T  Map results  Discussion and Conclusion

3. Motivation > a changing climate  More intense, frequent precipitation and flood events  Statistically show: very rare events become rare events  Precipitation sees a seasonal shift from summer to winter  Shift of the 0°C line > decisive for flooding in spring / autumn

4. Motivation > damage reduction  Improve preparedness before floods, enhance monitoring  More assets and values lie in flood-prone areas  Increase of risks and damages (2005: 3 Mia CHF)

5. User Needs & Objectives RADAR TEMPERATURE DISTRIBUTION EXTRACTION OF 0°C ISOTHERM TEMPERATURE POINT GAUGES INTERSECTION - Visual enhancements - Show attributes - toggle views BETTER ASSESSMENTS of catchment‘s disposition to flooding Real-time generation

6. Precipitation Radar and Temperature Interpolation  Radar today > integrated, multi-parameter, quantitative  Difficulties: instrumental, meterological factors affect accuracy  main advantage: spatial extent of prec. fields clearly visible  Temperature data > often inavailable in higher altitudes  Difficulties: interpolation accuracy in mountainous topography 1h data ≠ 1day data  spatial variability depends on temporal variability  main advantage: altitude is the main distribution factor

7. Precipitation radar maps > existing examples  Radar > from stand-alone in the 1960s to user-oriented quantitative monitoring products, storm-tracking, now-casting  Radar > uncertainties due to instrumental and meteorological factors  Radar > main advantage: spatial extent of precipitation field  Temperature > accuracy of interpolation depending on observation accuracy, point density and  Discussion and Conclusion No quantitative color scheme Too many classes Too coarse Way too many classes

8. Visual Improvements Radar  Radar > continuous, quantitative data [mm] or [in]  Reduce number of data classes  Use sequential color scheme, vary lightness  Apply visual smoothing for more genuine representations

9. Temperature maps > existing examples No legend, no clear allocation No areal interpolation

10. Visual Improvements Temperature  Temperature > continuous quantitative data [°C] or [°K]  Use diverging color schemes  Contrast hue, vary lightness for + and - values  Use point symbolizations AND interpolated surfaces AND extracted isolines

11. Taking advantages of web-mapping  …to avoid representational conflicts radar vs. temperature  Web-maps > Data exploration with interactive methods!  Web-maps > central calculations, visualizations on the client

12. Methodology > real time workflow radar

13. Methodology > real time workflow temperature

14. Interpolated temperature surface - Display of legend on mouseover - Display of ommited gauges

15. Temperature surface + framed rectangles - Display of time series, attributes on click - Red and blue rectangles on gauge sites

16. Interactive, radar image - re-classifed, re-colored, bilinear smoothing - Legend directly displayed in ‚raster‘ tab

17. smoothed radar image + 0°C isotherm -highlighting of area above 0°C - attributes directly displayed in ‚vector‘ tab

18. Framed rectangles for point temperature data - tooltip function on mouseover - attributes and legends directly displayed in ‚vector‘ tab

19. Discussion  Visual problems:  Complex workflows  exception handling  Other ways of handling missing/faulty data?  Data problems:  Other interpolation methods?  Calculation of real-time environmental lapse rate?  Inclusion of longitudinal lapse rate? Solar radiance?

20. Conclusion  Visual Improvements of real time radar possible in real-time! (inappropriate class numbers, illegible coloring, coarse resolution data)  Visual improvements of point temperature data (framed rectangles)  Real-time interpolation of temperature points (iso-line and statistical surface)  Distribution of maps over the web (Combined views, interactive exploration methods, remote assessment)

21. Thank you for your attention! Christophe Lienert, ETH Zurich, lienertc@ethz.ch http://RETICAH.ethz.ch