Usecase #1 - SST Target Data Set Evaluation
Version: - 1.0
A model developer would like to evaluate his/her numerical model of the North Atlantic by comparing existing derived sea surface temperature (SST) products with model generated SST fields: daily SST fields at 4km resolution for the period 1 January 2005 - 31 December 2006. The modeler knows that there exists many SST products, but s/he does not know where they exist nor how to access them. This use case is designed to provide the modeler with a tool that will allow him/her to find existing SST products, access them, match them with his/her fields and statistically compare the match-ups.
To begin with the modeler uses a search engine to find comparison SST datasets based on a suite of constraints that s/he provides as part of the search process: spatial/temporal coverage, spatial/temporal resolution, data origin (satellite, in situ or other model output), data type (climatology or time series), data access method. The result of this step is a list of candidate data sets.
The next step is to refine the list (if necessary) based on the metadata descriptions of the data. This step, at least at present, requires human intervention because the metadata associated with SST data sets is often not sufficiently complete to totally automate the procedure and because the number and character of discovered data sets is not known ahead of time. For example, the list may consist of a large number of qualitatively similar satellite-derived SST data sets which, if all were used, would overwhelm the computational capability of the modeler. The result of this step is the list of comparison data sets to be used in the analysis.
Comparison data sets fall into two classes: those comprised of point (a single lat, lon, time location - XBTs) and line (either a time series at a number of locations - mooring data - or lines in space-time - drifters or ship tracks) values in space and time and those consisting of 3D (lat, lon, time - satellite-derived data sets or model outputs) SST arrays. The modeler's data set is a 3D data set. For subsequent analysis, the modeler wants to match the observed values with values from the target data set based on a space-time window. The result of this step is a set of match-up elements. Match-up elements consist of SST values from the target data set, SST value(s) from a comparison data set and metadata characterizing the element.
In the next step, elements of the match-up data set are used to statistically compare the target dataset with the comparison datasets. The modeler would like to be able to experiment with this comparison, to perform it as a function of time of year, location, comparison data set, etc. S/he would also like to use a variety of commonly used statistical tools and a variety of commonly used visualization tools.
Finally, the modeler may want to perform a meta analysis of the data - an analysis across comparison data sets.
The user is likely to iterate on step(s) 1, 3 (and/or 4).
Author and date - SST Use Case Development Group 25 July 2007.
This material is based upon work supported by the National Science Foundation under award 0619060. Any opinions, findings and conclusions or recomendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).