Exercise Title: Converting a Data
Table to a Point Shape in Saga
Abstract: In this
exercise you will learn how to convert the data tables from ODV into
"point shapes" in Saga. In the process, the tables will be
inspected visually, and their histograms will be examined to find
possible valid data ranges versus possible invalid data ranges (for
temperatures in this example).
Point shapes can be displayed as data products in their own right, but we
intend to grid these data later.
Preliminary Reading (in
OceanTeacher, unless otherwise indicated):
|1. Use any good
ASCII editor, like Notepad++, ConTEXT, but
never Excel, to remove the extra header lines from all of the above
data subsets (the TXT files) and then save them. Currently there are
eleven of these unwanted lines on all ODV-derived spreadsheets.
You can add "noheader" to the filenames to help you remember these are
edited versions, if you want.
|2. Run Saga.
Then use FILE > TABLE > LOAD to load all 4 of the "noheader" tables.
|3. You should see these
objects in Saga now. Right-click on the first table, and select SHOW.
|4. It's always a good idea to
visually inspect data before use.
- Search the table to see what exported from ODV, and if there are any
empty fields (there should be none).
- Note that the date/time field is sometimes not complete or usable, a
You can close the table.
|5. The first thing we need to
do, is to convert this table to a point shape.
TOOLS > SHAPES-POINTS > CONVERT A TABLE TO POINTS
|6. This window appears.
|7. Make these choices:
- SHAPES > POINTS > CREATE - Saga recalls the last
job you did, and sometimes places the name of an existing file here.
TABLE, drop down the menu and select the first of the 4 tables.
NOTE: The fifth object, the world borders shape, is listed here
because it contains a table in its DBF file.
|8. For X choose the
LONGITUDE, and for Y choose the LATITUDE.
Then click OK.
Leave the Z line untouched to convert the entire data table to a
|9. This new object, a POINT
SHAPE appears in the data tree of Saga.
|10. Repeat this process with
the other 3 tables. But make sure that the top line always says
CREATE, when you run this module.
|11. You should have these 4
point shapes, when finished.
|12. Now, we should inspect
the point shapes. Right-click on the first one,
ADD TO MAP.
|13. Select NEW, and OK.
|14. Here are the points,
displayed using default colors. The JFM points are on the left, and
the 3500-4500 m products are on the bottom.
|15. Select any one of the
point shapes (in the left-hand menu) as the active object. In the properties panel
for this object,
make these selections:
- DISPLAY > OUTLINE - Uncheck the box
- COLORS > TYPE > Graduated color
- COLORS > SCALING > 100 COLORS
- COLORS > SCALING > PRESETS > Rainbow
- COLORS > VALUE RANGE > ATTRIBUTE > Temperature
Then click SETTINGS > APPLY
|16. Apply these
settings to all four maps. Notice that the VALUE RANGE is not set
above. Let Saga apply the natural range of each dataset, for now.
|17. Here are all four maps of
temperature values, organized as above. There's a lot of variation and
many color palette issues to be solved here, as you can see from the
There clearly is a cool area to the southwest in the deeper dataset.
|18. Right-click on each point
shape, and select SHOW HISTOGRAM to see that the distribution of values.
Here are all 4 graphs showing that there are many "outlier" values that need
to be evaluated.
Are they good, or should they be ignored in the analyses?
There appear to be 2 separate recommended ranges:
- 0-m Temp: 21-30 degrees C
- 4000-m Temp: 2.2-2.5 degrees C
Or something like that.
So you could change the value ranges to be limited to these extremes, and
perhaps your figures would look better. It's your data and your call.
|20. Now it's time to save
your work. Right-click on the first of the point shapes, and
select SAVE AS.
|21. Navigate to
the folder PRODUCTS > SAGA > VECTORS and save the point shape with the name
data_from_osd_jfm_3500m_4500m_liberia_wod_odv_saga. Saga will
actually save 5 files, with extensions SHP, DBF, PRJ, SHX and MSHP (metadata for
|22. In the same
way, save all your new point shapes, using these filenames:
|23. Now you have
the point shapes needed for gridding. All GIS software goes through
approximately these same steps, so the Saga
model is good training that should prepare you for other systems.