A NASA Climate Model Data Services (CDS) End-to-End System to Support Reanalysis Intercomparison

Friday, 19 December 2014
Laura Carriere1, Gerald L Potter2, Mark McInerney1, Denis Nadeau1, Yingshuo Shen1, Daniel Duffy3, John L Schnase2, Thomas Patrick Maxwell1 and Elisabeth Huffer2, (1)Climate Model Data Services, NASA Goddard Space Flight Center, Greenbelt, MD, United States, (2)NASA Goddard Space Flight Center, Greenbelt, MD, United States, (3)NASA Center for Climate Simulation, Greenbelt, MD, United States
The NASA Climate Model Data Service (CDS) and the NASA Center for Climate Simulation (NCCS) are collaborating to provide an end-to-end system for the comparative study of the major Reanalysis projects, currently, ECMWF ERA-Interim, NASA/GMAO MERRA, NOAA/NCEP CFSR, NOAA/ESRL 20CR, and JMA JRA25. Components of the system include the full spectrum of Climate Model Data Services; Data, Compute Services, Data Services, Analytic Services and Knowledge Services. The Data includes standard Reanalysis model output, and will be expanded to include gridded observations, and gridded Innovations (O-A and O-F). The NCCS High Performance Science Cloud provides the compute environment (storage, servers, and network). Data Services are provided through an Earth System Grid Federation (ESGF) data node complete with Live Access Server (LAS), Web Map Service (WMS) and Ultrascale Visualization Climate Data Analysis Tools (UV-CDAT) for visualization, as well as a collaborative interface through the Earth System CoG. Analytic Services include UV-CDAT for analysis and MERRA/AS, accessed via the CDS API, for computation services, both part of the CDS Climate Analytics as a Service (CAaaS). Knowledge Services include access to an Ontology browser, ODISEES, for metadata search and data retrieval. The result is a system that provides the ability for both reanalysis scientists and those scientists in need of reanalysis output to identify the data of interest, compare, compute, visualize, and research without the need for transferring large volumes of data, performing time consuming format conversions, and writing code for frequently run computations and visualizations.