Antarctic Temperature Approximating (WEEK 7)

As was discussed in the week 4 blog post, I suspect that warmer temperatures from the surface (lower portions of the atmosphere) directly above the Amery Ice Shelf contribute to weaker ice, and thus quicker rates of rift propagation. Luckily, unlike with the ocean temperature and sea ice concentration data, I did not have to download massive NetCDF files and extract data from them using Matlab code. I will discuss the approach I used below.

 

Surface temperature data: Spatial considerations

Unlike the SIC and ocean temperature data sets – where percentages or temperatures are given for rather large geographic areas with error ranges – the atmospheric temperature data are likely more reliable. There are numerous stations throughout Antarctica that measure and collect the temperature of the surrounding air at the same latitude and longitude each time (aside from potential minor tectonic shift).

While it’s helpful to know the exact location of where each temperature is being collected from, there is no temperature station on, or right beside, the Amery Ice Shelf. Thus, to estimate the temperatures over time, I determined which temperature stations are geographically closest to the AIS.

Figure 1: A list of temperature-monitoring stations in Antarctica, with distances from the AIS

Figure 1: A list of temperature-monitoring stations in Antarctica, with distances from the AIS

As can be seen from figure 1, the closest stations I found are as follows:

  • Zhongshan: 232 kilometers away from the AIS
    • This was the second research station in Antarctica established by China (the first being the Great Wall Station), and it was set up in 1989.
  • Davis: 253 kilometers away from the AIS
    • This station was named after captain John King Davis, an Australian explorer, and it was one of three set up by the Australian Antarctic Division, in 1957.
  • Mawson: 386 kilometers away from the AIS
    • This station was named after explorer Sir Douglas Mawson, and it was one of three set up by the Australian Antarctic Division, in 1954.

Once I found the closest temperature stations to the AIS, I next needed to figure out what direction they are in relative to the AIS. To do so, I used Google Earth Pro to plot the locations of the three stations, as well as the AIS. I realized that Mawson is west-northwest of the AIS, Zhongshan is east of the AIS, and Davis is northeast of the AIS.

Figure 2: Map showing distances and directions from temperature-monitoring stations to the AIS

Figure 2: Map showing distances and directions from temperature-monitoring stations to the AIS

Using these to predict the temperature of the AIS at any given moment, I assumed that a temperature gradient exists between the Mawson and Davis stations. In other words, if a straight line were to be drawn from the Mawson to Davis station, I assumed that the AIS temperatures would be somewhere between the two – and because the AIS is about three times closer to the  station, each AIS temperature would be more similar to the Mawson station than the Zhongshan station. I ended up not using the Davis station at all in these estimations. While this is not a perfect way to estimate the temperature, it could be more accurate than just using temperatures from one station. See figure 3 for approximations from 1999 and most of 2000.

Figure 3: Temperature approximations for the AIS, based on temperatures at Zhongshan and Mawson stations

Figure 3: Temperature approximations for the AIS, based on temperatures at Zhongshan and Mawson stations

 

The available atmospheric temperature data: Temporal considerations

 

The surface temperatures were provided as monthly averages. While this temporal resolution is much better than the 3-month resolution of the ocean temperature data, it is not as good as the daily resolution of the SIC data.

 

Also, not all temperature stations have an available temperature average for every month. As such, when using the station-distance method to approximate the AIS temperatures, there were several months that I was unable to use because either one or both of the stations would be missing data.

Figure 4: Examples of missing surface temperature data

Figure 4: Examples of missing surface temperature data

 

Even with the above considerations and constraints, I was still able to compile a decent amount of surface temperature data that, in the future, will be compared against AIS rift propagation rates to see if any relationship exists.

 

 

Speak Your Mind

*