U.S. National Science Foundation Project OCE 00-03970

Click on a logo below to visit a project site.

GIS, database design and web interfacing by Jeremey Bartley and Grimay Misgna

GIS database development and web interfacing by Casey McLaughlin

Database administrative support: Kurt Look

Correlations


Pairwise analysis of typology variables.

Coastal Cells

Microsoft Excel 2000
Comma Delimited Text file (.csv)
MS Excel 2000 (.zip)

Terrestrial Cells

Microsoft Excel 2000
Comma Delimited Text file (.csv)

MS Excel 2000 (.zip)

 


 

Budget comparison metadata

Dennis Swaney

September 2, 2001

 

Among other things, Vilma and I have revised and updated the LOICZ budget dataset during this week’s Lawrence, Kansas, meeting.  Below, we review the details of the  update, including the current list of variables, and the relationship of these data to corresponding typology data.

 

Several files have been generated (table 1), including

·        a comparison spreadsheet for all budgets

·        a comparison spreadsheet for those budgets with annual estimates

·        spreadsheets for both budget variables and associated typology variables

·        statistical summaries

 

The current version of the budget dataset includes 171 budget sites, including those from the Athens workshop.  The initial aim was to add a column for Vx for each budget because calculating this from the other data in the table proves to be inaccurate for multibox or multilayer systems.  We added this information (and in the case of stratified systems, added Vdeep, verified that the Vr term is as described in the LOICZ methodology pages). 

 

Separate entries for seasonal components have been added for all budgets where these data are available – the number of seasons vary with site.  The resulting spreadsheet contains in the neighborhood of 300 entries. A separate version of the dataset including only annual values has also been generated (entries may either result from annual calculations, or seasonally-weighted averages of budgets).  Because some budget sites only have data for a single sample or season, not all sites have an annual budget.  Some budgeted sites have multiple versions of annual budgets (e.g., the major Baltic components, or the Thukela river budget, which makes alternate assumptions about stoichiometry).  Some budgeted sites with more than one version on the web only have one entry in the database (in some cases, earlier versions have been updated; we may wish to review which sites, if any, maintain multiple entries in the comparison spreadsheet).

 

Multi-box systems

 

Boundary fluxes (VQ) are summed over all boxes.  VR and VX are reported for the outermost box  (oceanic boundary).

System salinity reported is that of the outermost box, seasonally weighted if necessary. System metabolism and “internal fluxes” are system totals, expressed per area of the total system.

 

Table 1.  LOICZ budget comparison data and statistical files

filename

File format

Description

Approx size (Kb)

variableunitsbudget&typologyLawrenceAugust2001.xls

excel

List of variable names in the comparison spreadsheet

60

annualbud&typolstats.XLS

excel

Statistical summary of annual budgets and corresponding typology variables

59

correlationsannbud&typol.XLS

excel

Correlation matrix (pairwise) for annual budget variables and corresponding typology variables

216

budgetcomparisonspreadsheetAug2001wseasonsVALUES.xls

excel

Full version of the comparison spreadsheet, including budget URLs –includes values but not formulas

972

budgetcomparisonspreadsheetAug2001wseasons.xls

excel

Full version of the comparison spreadsheet, including budget URLs –includes formulas (caution: sorting this spreadsheet may induce errors )

1150

budgetcomparisonspreadsheetAug2001annualVALUES.xls

excel

The comparison spreadsheet which contains only those sites with annual estimates of budget variables –includes values but not formulas

734

budgeuploadAug2001annualVALUESwodupes.xls

excel

comparison spreadsheet which contains only those sites with annual estimates of budget variables –with duplicate sites removed (e.g. multiple analyses for the Baltic, Thukela estuary etc)

610

budgeuploadAug2001annualVALUES.xls

excel

comparison spreadsheet which contains only those sites with annual estimates of budget variables –for uploading. (This includes some duplicate analyses for some budget sites)

631

budgeuploadAug2001annualVALUES.csv

Comma-delimited

Comma delimited version of the previous file

76

budavguploadAug2001annualVALUESandtypo.xls

excel

comparison spreadsheet which contains annual estimates of budget variables and corresponding typology variables averaged over the typology cells–for uploading. (i.e. corresponds to budgeuploadAug2001annualVALUES.xls  with typology)

946

budavguploadAug2001annualVALUESandtypo.csv

Comma- delimited

Comma delimited version of the previous file

222

 


Stratified systems

 

According to LOICZ convention, these systems do not have a Vx term.  Instead, the oceanic boundary is balanced by a deep flux, Vd  (generally into the system) and a surface flux (generally out of the system), the sum of -Vd and Vr. For these systems, Vd is reported in place of Vx.  The oceanic salinity reported is that associated with Vd (the deep water ocean salinity); system salinity is that associated with Vsurf, i.e. the salinity of the surface layer

 

Seasonal systems

 

In the absence of further complicating factors, annualized values for seasonal systems are calculated as the seasonally-weighted average of seasonal values.  If each season has equal duration, this is equivalent to a simple arithmetic average.

 

Variables in the dataset

 

A complete list of the variables, their definitions and units are provided in the file variableunitsbudget&typologyLawrenceAugust2001.xls.  The variables include both budget dataset variables and typology dataset variables.  Values of the typology dataset variables are calculated as follows:  for each budget site, the typology cells which coincide with it are grouped, and the averages and standard deviations over the group are calculated for each of the variables.  The label corresponding to the averages and standard deviations consist of the variable name with the prefix of ‘AV_’ or ‘SD_’, respectively.  Complete documentation for the typology variables is available at the website: http:www.kgs.ukans.edu/Hexacoral/Envirodata/

 

Statistical summaries

 

In addition to the files containing the budget data and the corresponding typology variables, we have created files with statistical summaries of these data and a pairwise correlation matrix (as we have also done for the typology database, described elsewhere).  Statistical information for each variable in the population of annual budgets is provided in table 2.

Table2.  Statistics calculated for each annual budget variable (and corresponding typology variables for the budget site)

Valid N

Mean

Median

Minimum

Maximum

Lower quartile

Upper quartile

Quartile range (= upper-lower quartile)

Variance

Standard deviation

Skewness (symmetric distribution=0)

Kurtosis (normal=0)

 

 

References

 

StatSoft, Inc. (1999). STATISTICA for Windows [Computer program manual]. Tulsa, OK: StatSoft, Inc., 2300 East 14th Street, Tulsa, OK 74104, phone: (918) 749-1119, fax: (918) 749-2217, email: info@statsoft.com, WEB: http://www.statsoft.com