Authors: Kevin Koy, Executive Director; Brian Galey, Web Application Developer; Matea Marsic, Web Application Developer; Maggi Kelly, Faculty Director, Geospatial Innovation Facility, Univ. California, Berkeley
The Cal-Adapt web application was launched in June of 2011 after nearly a year of planning, development, and testing. The site is designed to increase the availability and usefulness of modeled climate change data being produced by the scientific community in California. Through a combination of locally relevant information, visual¬ization tools, and access to primary data, Cal-Adapt allows a variety of different stakeholders access to locally relevant data that may better inform adaptation planning and decision making.
Cal-Adapt was developed by UC Berkeley’s Geospatial Innovation Facility (GIF), however all of the data and infor¬mation populating the site have been contributed from the Public Interest Energy Research (PIER) program’s vast network of research centers and facilities around the state. Support and guidance of the development process continues to be strongly linked with PIER leadership and the scientific community.
In the months following the site’s initial release, our team has been busy integrating new features, data, and tools in response to feedback and requests from the user community. This article highlights some of these new developments that are now available for use on the site.
NEW TOOL: EXTREME HEAT
One of the most serious threats to the public health of Californians, that has already presented a challenge to date, are extreme heat events. Climate models, developed by the Scripps Institution of Oceanography, project that extreme heat events will increase in frequency, intensity, and duration given future climate change.
Cal-Adapt features several different tools to visualize the projected increase in temperatures. However, at the time of the site’s initial launch these tools were limited to using only monthly temperature data. In order to visualize and under¬stand extreme heat events, one must have access to projected temperatures at a daily time step. There are several thousand raster data layers per scenario and model at this temporal frequency. This enormous amount of data presents a process¬ing challenge to provide a fast and dynamic visualization tool that can effectively describe the projected patterns of extreme heat events at a given location.
In order to address this need, the Cal-Adapt development team utilizes a powerful and flexible open source stack includ¬ing the Python web framework Django, the cartographic styling capabilities of MapServer, and some essential Python packages for scientific computing, including NumPy which provides a powerful, multidimensional array objects for quick resampling of time series data. The user interface relies on jQuery with jQueryUI, and takes advantage of the HTML5 canvas element for the more interactive graph tools.
This setup allows for rapid access to the data that can then be used to populate a variety of charts and tables. Upon entering the tool the user is instructed to select a location on the map, and then is led through a series of visualizations that show:
a. Number of extreme heat days by year: The number of projected extreme heat days per year (1950-2099)
b. Number of warm nights by year: The number of project¬ed warm nights per year (1950-2099)
c. Number of heat waves: The projected total number of heat waves (1950-2099)
d. Timing of extreme heat days by year: The projected timing of extreme heat days from April through October (1950-2099)
e. Maximum duration of heat waves by year: The number of days in the longest heat wave projected by year (1950-2099)
f. Daily high temperatures by year: The projected daily temperature between April and October (1950-2099)
The available charts and graphics for this tool are the same as those used by scientists to describe the potential increases in extreme heat events, however these graphics are usually limited to one or two locations. By integrating the data into a dynamic online mapping environment, users can now access this same information for any location throughout the state.
IMPROVED DISPLAY OF UNCERTAINTY
Improving on the public’s understanding of the uncertainty in projection data was a high priority for development as much of the feedback we received often revolved around this subject. There are a variety of different models and scenarios for any given type of data (temperature, precipitation, etc.), and it is important to convey the fact that one line in a chart only represents one of many potential outcomes. Given the fact that it is impossible to predict exactly what the future will look like, particularly in terms of demographics, socioeconomic factors, and scientific innovations, many different projec¬tions are made in an attempt to cover the range of potential outcomes that may be expected.
FIGURE 2 – Monthly temperature over time for Oakland, CA (a), and monthly temperature over time including range of model results (b)
Reprinted with Permission, Bay Area Automated Mapping Association