Hazus Users Conference 2013

5-7 August 2013.

Dave, Kevin, Christina and Brendan were breathing sighs of relief on Monday morning as the delegates to the 2013 Hazus Users Conference started showing up at the reception desk. With Eric Berman, they greeted friends and colleagues congregating at the government centre in Indianapolis to share their stories about doing Hazus and ways to make it better. Eventually over a hundred people showed up, mostly from the eastern US. The rest came from further afield in the US, Canada, and Singapore.  The meeting was different than 3 years ago in Indianapolis. The stories and discussion were deeper, more technical and philosophical, more urgent. That was reflected by a suggestion that the conference include peer-reviewed papers. I fretted that my presentation on using Hazus for Canadian earthquake response exercises was too old school and of little value to the sophistication of the discussion. It was good to have focused on how the exercise teams used the Hazus disaster scenarios rather than on the scenarios (standard Hazus outputs). On Wednesday morning we were wrapping up the meeting with a 2-hour discussion on improving the Hazus system, and evolving with present and future needs. Kevin built anticipation for the discussion by handing out a five-question form on Monday (homework). All the slide decks from the conference were collected and distributed to participants at the end of the meeting. I got excited about John Remo's flood evaluation for Illinois, Heidi Crow's presentations testing agricultural loss estimates of Hazus floods, Raju's talk on using Hazus to measure potential losses from climate change induced sea level rise and coastal flooding in Singapore, and the discussion scattered between talks about the best way to do a flood-loss estimate. Thanks very much to Atkins and Arc Aspicio for the food and drink. It was so nice not to have to go scrounging through the government building for a coffee and snack.

The MCs generously highlighted the advances being made in Canada for using Hazus. Miro Nastev excellently summarized Canadian Hazus and related work. Organizers set up a wonderful private reception on Monday evening at the nearby historical museum, where we could enjoy conversation, hors d'oeuvres and an exhibition on the massive Indianapolis flood of a hundered years ago, 1913.

John's flood evaluation for Illinois incorporated CommunityViz to visualize the Hazus outputs and used aggregated loss indicators much in the way Murray Journeay, Geological Survey of Canada, developed for use in Canada. They ought to get together and compare notes (maybe they already did).

Heidi and her Masters thesis supervisor, Jordan Hastings, USC, discussed limitations and a mistake they discovered in the crop loss-estimation tool of the Hazus flood module. Heidi made a case for including agricultural growth zones. They demonstrated it was important to have regional calendar growth curves that can be shifted dependent on the years and to include the effect of changes in growth patterns due to local elevation. She made a case for including the flood water depth compared to the growth stage. Young plants will drown within a day in shallow water and yet as mature plants will survive longer in deeper water. They showed that Hazus predicts almost double the actual flooded crop areas, because of a clipping error of agricultural polygons.  She was invited to give her presentation to the Canadian Risk and Hazards Network Symposium (CRHNet) in Regina, Saskatchewan, November 5-8, 2013 in the special session where Hazus and LIRA, the Agriculture Canada flood crop loss tool will be discussed.

Dr. Durairaju Kumaran Raju made a clear, well presented case for the value of using loss-estimation to evaluate the risk to Singapore of sea-level rise and storm-surge caused by a warming and stormier climate. Working with the Polis Centre in Indianapolis Dr. Raju's team set up Singapore inventory in Hazus as an Hawaiian island and could conduct a full user defined facility coastal flood analysis. They built the detailed inventory for structures in an area pre-determined to be susceptible to sea-level flooding predicted by the Intergovernmental Panel on Climate Change (IPCC). We really have to write the manual on how to trick Hazus for use anywhere in the world.

Hazus floods overestimates losses for areas where aggregated data in census tracts and blocks intersects but a small portion of the flood plane. Several solutions were proposed, some of which have been proposed for some time.  Nikolai Todorov explained the dasymmetric approach, which used an intersection of the census block data with that of land cover datasets that classify  polygons by the density of population. Applying weighting factors to the census block aggregated data from the ratios of the land-cover densities better approximates potential flood losses. Lots of debate about this approach, perhaps because it still is an approximation and relies on how well the weighting factors are determined.

Lots of ideas and discussion about fixing, improving and morphing Hazus; too much for this report. In summary people wanted the bugs fixed now, adapted to work with ArcMap 10.2 now, disengaged, and not divorced, from Arc in the future and moved to open source and online. Mixed reviews on what to do about access to consolidated Hazus-compatible assett inventory.

Plans are in the works to hold a Hazus Users Conference in August of 2014, probably in Indianapolis, unless another group steps up now to take it on. Too bad its complicated for the American contingent to cross the border for such a conference. Ottawa or Vancouver would be a great place, as would Singapore!!