CDIAC Ameri. Flux Data Management Tom Boden (Bai Yang) Carbon Dioxide Information Analysis Center (CDIAC) Oak Ridge National Laboratory (ORNL) Presented September 27, 2010 at the CDIAC User Working Group Meeting Oak Ridge National Laboratory OAK RIDGE NATIONAL LABORATORY U. S. DEPARTMENT OF ENERGY
Ameri. Flux Network ~140 sites in 5 countries; 92 active sites, 48 inactive sites Participation Requirements Make year-round core measurements using the eddycovariance technique Submit data to CDIAC within 1 year of collection Participate in Ameri. Flux Science Meetings and synthesis/modeling activities DOE supports the Ameri. Flux infrastructure including the chairperson, calibration laboratory, and a permanent data archive (CDIAC since 1997). 2
Global Network Regional Networks FLUXNET - a network of networks Ameri. Flux Canadian CP Americas CARBOEUROFLUX Asia. Flux Oz. Net Europe Asia - Japan Oceania Nonnetwork sites Tower Sites Architecture of Global / Regional Flux Networks
Primary FY 2010 CDIAC Ameri. Flux Team Misha Krassovski – data system - (0. 9 FTE) Bai Yang – micrometeorologist [QA/QC] - (0. 8 FTE) Barbara Jackson – SAS programming - (0. 7 FTE) Ben Norton – web & DB maintenance - (0. 5 FTE) Tom Boden – coordination - (0. 3 FTE) Lianhong Gu – A-Ci analysis – (0. 04 FTE)
Ameri. Flux Data Processing & Products Site Investigator & Team Flux/Met Data Biological Data Level 0 Site Proc. CDIAC QA/QC Network-wide 30 min Database Averages Standard Files BADM Level 1 FLUXNET Quality Flags Gap-filled NEP/Re CADM Level 2 Level 3 & 4
Level 1 to Level 2 Data Processing Scheme Level 1 data received from Ameri. Flux site PIs Flux, Meteorological, and Ecological Data Files • • Stored “as received” from investigator File format, parameter names, and units unchanged Add value through automated QA/QC Perform QA/QC Calculate additional parameters, add to network database Incorporate into Web data interface & produce standardized data files & metadata reports • • • Range checks, missing values, check solar elevation and time stamp Basic Statistics (mean, variance, range) Gap Analysis: determine frequency and duration of gaps in record Spike detection, stationarity, diurnal and seasonal pattern Cross checks/correlations (Rg vs. PAR) Graphical Analysis: time series, property vs property, frequency Incorporate into network-wide database • Convert into standard Ameri. Flux units, reporting intervals, and naming conventions • Generate core parameters (e. g. , VPD) where needed Web access to network-wide database • • Querying capabilities – time, measurement parameter, and site Graphics generation and data retrieval capabilities FTP access to standard files • • Standard files with 40 core variables (csv, Net. CDF, tar) Metadata files (html & txt formats) L 1 files available from FTP area | | 5 days | | L 2 data products available via Web interface and from FTP area 6
Progress in gap-filling meteorological data · Completed. We have completed a prototype data set of gap-filled meteorological data for the majority of active Ameri. Flux sites. · Ongoing. This data set has been evaluated against other data sets filled by different techniques. The evaluation will continue if necessary and the sensitivity of ecosystem models to the gap-filled data sets will be estimated. · Ongoing. We are improving our gap-filling methods by including reanalysis data sets (NARR, Day. Met, etc) and radar precipitation data. Present focus on filling long gaps and improving data reliability for precipitation, radiation, soil moisture and soil temperature. · Goal. We hope to finalize the gap-filled met-data sets before the next Ameri. Flux annual meeting (Feb. 2011).
Ameri. Flux Data System http: //ameriflux. ornl. gov Metadata Site & PI Information, Pubs, & Instruments (e. g. , Fc corrections, instrument calibrations FY 2011 Users Worldwide PHP Interface Flux & Meteorological Data Commenced 8/2009 http: //ameriflux. ornl. gov/edit 12/2008 CDIAC MS SQL 2005 RDBMS Biological Data BADM Reported Variables Flux & Met/Bio 1/2009 3/2009 Leaf Aci ORNL Ameri. Flux & FLUXNET Data Team PHP Interface 3/2010 11/2009 FY 2010 Priorities Data extraction/e-mail notification Load remaining L 2 data, begin loading L 4 data Web migration Inclusion of L 2 & L 4 data in the Earth System Grid Biological data editing functionality http: //leafweb. ornl. gov Gu EDO methodology & code 8 Web interface/Sharepoint for data submission and dissemination of results
Examples of CDIAC Ameri. Flux Contributions to Synthesis & Modeling Activities · La Thuile, Italy workshop and resulting data sets · NACP Interim Site Synthesis · C-LAMP
La Thuile, Italy FLUXNET Workshop February 18 -22, 2007 Mandatory data: NEE/FC ~60 participants, good representation from regional CO 2/SFC U*/TAU networks and major programs (LBA, TCOS) Rg/PAR 921 site yrs of data from 240 sites worldwide (33 countries)Ta H 2 O/RH/VPD 318 site yrs of data from 77 Ameri. Flux sites · Participation and data response – – – · Resulted in ~60 proposed papers – http: //www. fluxdata. org · Workshop highlighted the importance of biological/ecological data and consistent ancillary information (ecosystem classifications)
C-LAMP The Carbon-Land Model Intercomparison Project (C-LAMP) consists of an experimental protocol, model evaluation metrics, a prototype diagnostics package, model output standards, and a database of model simulation results on the Earth System Grid (ESG). See http: //www. climatemodeling. org/c-lamp In the first set of runs, over 16 ky of simulation and 50 TB of output were generated using the Climate Science End Station INCITE allocation at ORNL. CLM 3. 1 (Community Land Model) combined with CASA´ (Carnegie-Ames-Stanford Approach) and CN (carbon & nitrogen) biogeochemistry modules were evaluated against best-available satellite- and ground-based measurements, and new runs are being used to evaluate CLM 4 performance. C-LAMP will serve as a benchmarking prototype for Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report (AR 5) in conjunction with ESG. Randerson, J. T. , F. M. Hoffman, P. E. Thornton, N. M. Mahowald, K. Lindsay, Y. H. Lee, C. D. Nevison, S. C. Doney, G. B. Bonan, R. Stöckli, C. C. Covey, S. W. Running, and I. Y. Fung. “Systematic Assessment of Terrestrial Biogeochemistry in Coupled Climate-Carbon Models. ” Global Change Biology, in press.
Ameri. Flux Data Are An Important Component of C-LAMP diagnostics include model comparisons with MODIS (Moderate Resolution Imaging Spectroradiometer) leaf area index (LAI) and net primary production (NPP), Globalview amplitude and phase, Ameri. Flux Level 4 energy and carbon fluxes, Free Air CO 2 Enrichment (FACE) site measurements, and various estimates of carbon stocks and transient dynamics. Shown at right is a time series comparison of net ecosystem exchange (NEE), incoming shortwave, latent and sensible heat, gross primary production (GPP), and ecosystem respiration from CLM 3. 1 -CN against measurements from the Morgan Monroe site. C-LAMP diagnostics must be run before any model changes to the CLM biogeophysics or biogeochemistry can be made.
Future Directions & Emphasis? Gap-filling meteorological data Biological data QA/QC of existing submissions Improved parameterizations for models Expanding Leaf. Web Amassing key measurements (e. g. , soil respiration) Flux re-analysis product within the ORNL SFA framework FY 2011 DOE BER “Ameri. Flux Center” solicitation Collaborations with NEON 13
Скачать презентацию CDIAC Ameri Flux Data Management Tom Boden Bai
56dc32338bc185289013ae7f25d2e0e2.ppt