Service Oriented Gridded Atmospheric Radiances (SOAR)

We are developing a scalable web service tool that can provide complex griding services on-demand for atmospheric radiance data sets from multiple temperature and moisture sounding sensors on the NASA and NOAA polar orbiting satellites collected over the past three decades. This server-to-server middle ware tool will provide the framework for transforming user requests for an arbitrary spatial/temporal/spectral gridded radiance data set from one or more instruments into an action to invoke a griding process from a set of scientifically validated application programs that have been developed to perform such functions. The invoked web service agents will access, subset, concatenate, convolve, perform statistical and physically based griding operations and present the data as specified level 3 gridded fields for analysis and visualization in multiple formats. Examples of the griding operations consist of spatial-temporal radiance averaging accounting for the field of view instrument response function, first footprint in grid bin, selecting min/max brightness temperatures within a grid element, ratios of channels, filtering, convolving high resolution spectral radiances to match broader band spectral radiances, limb adjustments, calculating variances of radiances falling in grid box and creating visual displays of these fields. The gridded web services tool will support both human input through a WWW GUI as well as a direct computer request through a W3C SOAP/XML web service interface. It will generate regional and global gridded data sets on demand. A second effort will demonstrate the ability to locate, access, subset and grid radiance data for any time period and resolution from remote archives of NOAA and NASA data. The system will queue the work flow requests, stage processing and delivery of arbitrary gridded data sets in a data base and notify the users when the request is completed. This tool will greatly expand satellite sounding data utilization by responsively meeting diverse user specified requests in terms of the spatial and temporal compositing of radiance fields. Moreover, the volume of sounder data records produced from current and future instruments varies from GB's to TB's per day and griding these sounding data can thin the volume to KB's to MB's per day making them easier to download to desktops and laptops. This not only will better serve a wider earth science community but makes these capabilities more readily useful to the education community. This presentation will describe the rationale for the project, an overview of the system architecture, a description of the framework for executing the applications on the distributed cluster and present examples of gridded service requests that are currently available. This demonstration project represents a foundation for the development of a distributed web service architecture that will be able to invoke requested services for temperature and moisture retrievals for arbitrary integrated gridded radiance data sets. We plan to extend the framework to accommodate such services for other earth observing instruments as well.