Data in Four Dimensions (WEEK 5)

As I mentioned in the previous blog post, I suspect that warmer ocean temperatures correlate with higher rates of rift propagation. However, throughout the Amery Ice Shelf, the depths of it range from roughly 300 meters at the calving front (point at the end of the AIS, where ice typically breaks off from) to 1,200 meters at the calving front (point at which the ice is no longer attached to the underlying bedrock). Because I do not know how thick the ice is that directly surrounds the rift, I compiled as many relevant depths as I could that could be in contact with that specific portion of the AIS.


The available ocean temperature data

Below is a rather complicated-looking snippet of the available ocean temperature data, as it is shown on Matlab:


Figure 1: Ocean temperature data from a netcdf file, displayed through Matlab

As can be seen, each datum displays one of three things: a non-zero number, a “0”, or “NaN”. The non-zero numbers represent temperature anomalies, which are calculated by subtracting out a uniformly standard mean value from the corresponding location- and time-specific observed temperature value. The zeros represent areas where temperature anomalies could have been calculated for that location and time, but were not. And the “NaN”s represent areas where no ocean temperature anomaly could be calculated, due to there being land or ice in that area.


What with this massive grid, it may seem challenging to sort through all the data. But as the title suggests, the ocean temperature data are available in four “dimensions”: latitude, longitude, depth, and time.

  • Latitude & longitude: the geographic range of each temperature (datum in figure 1) is one-degree latitude by one-degree longitude. That initially might not sound like a large size, but when converted into the kilometers in the area of the Amery Ice Shelf (so 68-69 degrees S and 71-72 degrees E), that translates to about 111 by 40 kilometers. Also note that the data sets used here are of the entire planet, so naturally it makes sense that they would encompass such a large area.
  • Depth: the depth intervals of the data are provided as follows. As can be seen, the intervals between consecutive depths vary, starting off as low as 10 meters and increasing to 250 meters at the bottom. Unfortunately though, the available data for the geographic area of the Amery Ice Shelf do not go lower than 700 meters. While I am not entirely sure why, this may be because the AIS is such a remote place that attempting to get an ocean temperature measurement of an area so entrenched by bedrock and ice would be extremely challenging.
Shallowest In-between Deepest
0 meters 200 meters 1,000 meters
10 meters 250 meters 1,100 meters
20 meters 300 meters 1,200 meters
30 meters 400 meters 1,300 meters
50 meters 500 meters 1,400 meters
75 meters 600 meters 1,500 meters
100 meters 700 meters 1,750 meters
125 meters 800 meters 2,000 meters
150 meters 900 meters
  • Time: the temperature data are provided in three-month ranges (January-March; April-June; July- September; and October-December). This is also not optimal, as they do span a relatively large time range.

Extracting ocean temperature data

To extract the ocean data at the latitude & longitude of interest, I downloaded all the netCDF files I could from 2001 to the present, and I used a special code in Matlab. Below is an example code segment that I used to extract data from January-March 2009, with a few annotations:

Figure 2: Matlab code used to extract ocean temperatures around the Amery Ice Shelf

Figure 2: Matlab code used to extract ocean temperatures around the Amery Ice Shelf

  • Line 1 of code imports the file of interest.
  • Line 2 of code specifies the spreadsheet of data that I want. In this case, I want “t___”, which is the {{standard unflagged …}}.
  • In the following lines, there are three sets of bracketed numbers.
    • The first number represents the rank of the longitude of interest; the longitude is not 253 degrees, but rather the 253rd value of longitude in this file is -68.5 degrees (which represents the area between -68 & -69 degrees longitude).
    • The second number represents the rank of the latitude of interest; the latitude is not 22 degrees, but rather the 22nd value of latitude in this file is 72.5 degrees (which represents the area between 72 & 73 degrees latitude).
    • The third number represents the rank of the depth of interest. In this file, [12] represents 275-350-meter depths; [13] represents 350-450-meter depths; [14] represents 450-550-meter depths; [15] represents 550-650-meter depths; and [16] represents 650-750-meter depths.
Figure 3: results from Matlab code that extracts ocean temperature data from the Amery Ice Shelf

Figure 3: results from Matlab code that extracts ocean temperature data from the Amery Ice Shelf


Compiling ocean temperature data

As Matlab provided me with the temperature values of interest, I copied and pasted them all onto an Excel spreadsheet.

Figure 4: Compilation of ocean temperatures from relevant dates

Figure 4: Compilation of ocean temperatures from relevant dates

I then began analyzing the data to find trends and patterns within it. However, rather than talk about some of those trends and findings, I will instead end this blog post here, and save the introduction to my next data set (sea ice concentration) for the following blog post.