DATE: 30 April / 1 May, 2001
Create as much space as needed within the form.
Geographic Zones Selected (leave blank if done by range):
Geographic Range Selected (leave blank if done by zones):
Mexico, including part of South USA and Central America. 0 to 40° Lat N; 60 to 130° Long. W.
Cell types selected from database (Terrestrial, Coastal, OceanicI, OceanicII, OceanicIII):
Coastal
Variables and ranges selected from database (if filtered – if not filtered, no entry is needed):
Variables selected:
Runoff, annual mean (mm/yr) [Continous]
Land Elev, std dev of G30 values [Continuous]
Cell Landcover, % Cropland
Variables added (if any):
What do you expect to get out of clustering these variables or what are you trying to achieve?
(enter name of document if experiment is written up separately
Save the Cluster summary – if will contain much of the information needed!
(Instructions for summary and data saving are at the end of this document)
Cluster Run Number:__
Variables selected in cluster (include any weights if not set at 1): 3
Number of Clusters: 10
Distance Method: [ x ]Average [ ]Maximum
COMMENTS OR DESCRIPTION ON CLUSTER OUTPUT:
1. Once we got acquainted with the databases, we defined that our target would be 1) to see if we could reproduced our available classification of the Mexican Coastal Zone with the typological tools available in the workshop, and 2) then apply our design to the global coastal zone to see if it was useful.
2. First we revised our available coastal classification (revised from Merino M.,1987, The Coastal Zone of Mexico, Coastal Management, 15: 27-42) to adequate it to the objectives of LOICZ. The main change was that we eliminated the different types or regions that were based mainly in oceanographic differences. As a result the Caribbean and the rest of the coast of the Yucatan Peninsula were considered a single region, and the Baja California Peninsula also became a single region.
3. As a second step we defined three groups of parameters or proxies that we though adequate to classify the coastal zone of Mexico. One group was chosen to represent coastal runoff, another for the geomorphic nature (i.e. existence or not of coastal plain and continental shelf ) of the coast and a third the relative magnitude of C,N,P inputs to the sea. The parameters we chose to try were:
As runoff proxies:
Precip, ann mean [Continuous]
Runoff, annual mean (mm/yr) [Continuous] (*Final selection)
Basin Runoff (1000s of cubic meters)
Soil Carbonate C, kgC/m2 (0-1m) [Discrete] (This was included to see if could help discriminate karstic areas where groundwater flow to the coast would be dominant –it was, however, useless-)
As geomorphic proxies:
CZ Bath/Elev, std dev of SS2 value
Ocean Bath, mean SS2 value [Continuous]
Land Elev, mean G30 value [Continuous]
Land Elev, std dev of G30 values [Continuous] (*Final selection)
Cell Landcover, % Water
Basin Landcover, % Water
As C, N, and P, flux proxies:
Soil Organic C, kgC/m2 (0-30cm) [Discrete]
Cell Landcover, % Cropland (*Final selection)
Cell Landcover, % Grassland
Cell Landcover, % Urban and Built-up
Population, 30' cell total [Continuous]
Basin Landcover, % Cropland
Basin Landcover, % Grassland
Basin Landcover, % Urban and Built-up
Basin Population
4. For our experiment we chose a region that extended widely around Mexico, in order to be sure to include the boundaries of the clusters formed, and also to make the exercise of wider application. The range selected (0 to 40 Lat N, and 60 to 130 Long W ) included all of the Caribbean region, Central America, the northern part of South America (Colombia and Venezuela) and Most of the USA.
5. We then tried individually each of the selected parameters to see how well they worked as proxies of each component in terms of dividing the coast in the regions defined initially. We then assigned the parameters within each group a priority order in terms of their effectiveness.
6. The three parameters selected (indicated by * in the list) were then used to create a coastal typology. The initial unweighted and unfiltered combination of “Runoff, annual mean (mm/yr)”, “Land Elev, std dev of G30 values” and “Cell Landcover, % Cropland” turned out to better represent the Mexican coast typology we used as a target. The file is named “P+ESD+%Cr-ww10” (we kept the original awful name to avoid problems with the link to the explored related files).
7. The addition of other parameters within the initial selection did not improve the ability to generate the target typology, nor did the weighting of the parameters.
8. Although parameters that were intended to account for coastal cities (like population and % Urban Landcover) were not useful in the case of Mexico, we believe this due to the absence of important coastal urban centers in this country, in contrast to other.
9. The combination so selected was then applied to the world coastline to see if it generated a useful typology. This experiment can be found in the file “World-Mexico”. We added to the three parameters used for Mexico, a selection one that discriminated between frozen (<1 C) and not frozen (>1 C) areas and SST Monthly mean.
Additional Comments on problems or improvements needed:
**************************************************************************************
1. % Soil Carbonate Content seems not to be useful for distinguishing Karstic regions, where ground water input would substitute surface runoff. The data are discrete, and relatively poor.
SAVE CLUSTER SUMMARY:
Netscape Navigator:
Step 1: Click on the Cluster Summary button
Step 2: From the “File” menu, select “Save As . . .”
Step 3: Create a new directory, and open that directory
Step 4: Give your file a useful name, and the extension .htm
Step 5: Click “Save”
Step 6: Right click on the image (the map)
Step 7: Select “Save Image As”
Step 8: Save the image in the same directory
Internet Explorer:
Step 1: Click the Cluster Summary button
Step 2: From the “File” menu, select “Save As . . .”
Step 3: Create a new directory, and open that directory
Step 4: Give your file a useful name, and click “Save”
Microsoft Word 2000:
Step 1: Click the Cluster Summary button
Step 2: From the “Edit” menu, click “Select All”
Step 3: From the “Edit” menu, click copy
Step 4: Open a new document in Word 2000
Step 5: From the “Edit” menu, select “Paste”
Step1: Click on the View image as one layer button
Step2: After the new window pops up Right-Click on Image and select the Save image as option
Step3: Place the image in the Loicz folder on your desktop
Insert Image into the space below:
NOTES ON SAVING YOUR INFORMATION (output data):
Cluster information:
Step1: On the Visualize tab click on “Source” button
Step2: Under the View Column click on the “clu” button
Step3: Select all of the text in the bottom right window and copy it to the clipboard (edit copy or ctrl-C)
Step4: Open Notepad and paste the information in and save it into your LOICZ folder
Step5: Save the file with a unique name like clusterNum_clu.txt and then write the name below:
Cluster assignment for each data cell:
ONLY DO THIS IF YOU PLAN ON USING IT IN ARCVIEW!!!
Step1: On the Visualize tab click on “Source” button
Step2: Under the View Column click on the “tag” button
Step3: Select all of the text in the bottom right window and copy it to the clipboard (edit copy or ctrl-C)
Step4: Open Notepad and paste the information in and save it into your LOICZ
Step5: Save the file with a unique name like clusterdata_tag.txt and then write the name below: