NASA Goddard GIOVANNI Support for HTAP

To address a fuller understanding of intercontinental transport of air pollution in the Northern Hemisphere, the Task Force on Hemispheric Transport of Air Pollution (HTAP) is conducting a series of multi-model evaluation and intercomparison experiments to: produce estimates of intercontinental source-receptor relationships; improve our understanding of the variability and uncertainty in current model estimates; assess potential future changes in source receptor relationships; and guide future model developments to decrease uncertainties in source-receptor relationships. A necessity to inter-compare data from multiple models provides various technical challenges. The existing HEMITAP, the specially developed IDL-based desktop tool, requires all of the data to be on the users’ computer, i.e., users have to replicate the HTAP archive on their local workstation or run analysis remotely on the HTAP server at Juelich, Germany, using very slow x-protocol. We present a new option to run analyses online using Giovanni, the NASA Goddard Interactive Online Visualization ANd aNalysis Infrastructure tool for visualization and analysis of very large, global datasets. We describe our approach to extend the Giovanni analytical and statistical tools to support HTAP. While current versions of Giovanni offer fast and easy access to a wealth of satellite remote sensing data along with a variety of visualization and analysis tools, the HTAP Giovanni concentrates on model data intercomparison and provides the capability to examine spatial variability and transport of air pollutants on regional and global scales. HTAP Giovanni contains most of the functionalities available in the existing desktop HemiTAP tool, but the implementation is web-based, so all of the visualization, analysis and data extraction may be performed online using a regular web browser, without any need to remotely log on to HTAP servers or download data and software to a local machine. The initial interface, which includes data from several of 40+ atmospheric chemical transport models participated in the HTAP project, is designed to let users easily navigate through multiple models, experiments, diagnostics, variables, and analysis options. The newly developed multi-faced user interface is comprised of related lists: the selection from one list determines the options to be presented in the next. This approach allows us to present all options and selections on a single page. Time series plots with statistics, scatter plots, and correlation coefficient and difference maps for models allow comparison between remote sensing data and models. Some of the original model data, available on sigma vertical coordinates, are interpolated at pre-defined pressure levels based on the reference pressure and surface pressure values provided by each model. Several issues with non-standard data have created challenges that are being addressed by all the parties. The HTAP at Juelich has recently adopted the Web Coverage Service (WCS) as data transfer protocol. The implementation of a set of interoperability standards is a necessary condition for ensuring interaction between data centers hosting different types of data. HTAP Giovanni plans to utilize this service.