How to Collect Data from 14,500km Away

One of the first steps to this research project is collecting data. Given the need to collect visual data from Antarctica, which is approximately 9,000 miles (or 14,5000 kilometers) away from Virginia, the best solution would be to go to Antarctica itself. However, because that is not feasible time- or money-wise, I am instead investigating the Amery Ice Shelf through satellite imagery: data that is remotely sensed by satellites circulating in high altitudes, then posted online as pixelated images.

Most of the images I am using come from the National Snow and Ice Data Center’s database. And the type of satellite being used for these images is MODIS: The Moderate Resolution Imaging Spectroradiometer. MODIS, along with some other satellites, use a variety of spectral bands in a variety of wavelengths. The images of the Amery Ice Shelf have pixel resolutions of 250-by-250 meters or 0.25-by-0.25 kilometers. In other words, every pixel represents a wavelength value averaged within an area of 62,500 square meters (or 0.0625 square kilometers).

Figure 1: The MODIS satellite circling Earth (source:

Figure 1: The MODIS satellite circling Earth (source:

I will also be using data from a second satellite: Landsat-8. Compared to MODIS, Landsat-8 has images of 30-by-30 meters, which gives much more precision than the 250-by-250-meter resolution of MODIS. However, Landsat-8 was not launched until February 2013, so as such, Landsat-8 does not provide satellite images from before 2013.



The most ideal way to study change in Antarctica through satellite imagery would be to compile as many satellite images as possible and measure them. Ideally, this would be one image from every day, or even every hour, or even every minute, etc. However, there are many limitations that prevent me from being able to collect this much data.

  1. Antarctic declination: This first limitation solely concerns a few areas of the planet. Antarctica is far enough south on the planet that it experiences drastic changes in daylight throughout the year. A such, the sun never sets during most of the Southern hemisphere’s summer (December-February), while the sun never rises – or at best appears in twilight phase – during most of the Southern hemisphere’s winter (June-August). As such, as is shown in figure 2, there are a lack of visible data from June-August.
  1. Obstructions: Because satellites circle the Earth in such high altitudes, it is entirely possible that there can be interferences between the satellite and surficial feature of interest. Clouds are a common example of an interference. Figure # shows an example of clouds preventing us from being able to measure the feature of interest.
  1. Available data: Even when there is light and a lack of clouds in Antarctica, there is still a limit to the frequency with which a satellite can take images. For example, in January 2013, the only images available are January 1st and 19th; none of the other days have any data associated with them.
  1. Researcher’s time: Even with the remaining images, there are still way more than I can feasibly compile and measure over the summer, since I am examining roughly two decades of data.
Figure 2: Available images from 2008 (source:

Figure 2: Available images from 2008 (source:

Despite the above limitations, I have done my best to compile at least two images from every year, ideally each landing between the Antarctic summer and Antarctic winter (so one from March-May, and the other from September-November). Figure 3 shows the number of satellite images from each year that I will be measuring.

Figure 3: Number of satellite images per year being measured

Figure 3: Number of satellite images per year being measured


In the next blog post, I will go into detail about the specific measurements of interest.

Speak Your Mind