Snow Cover Detection {Methodology Using Landsat}
NASA MODIS Land February 2004
One important use of snow cover analysis is for snowpack estimation. Mountain snowpack is a key component of the hydrologic cycle that stores water from the winter as snow and releases it as surface runoff and groundwater recharge in the spring and summer. The volume water of stored in a mountain range as snowpack is known as the snow water equivalent (SWE) and gives a good indication of the volume of runoff and therefore the water available for use later in the year. In some areas of North America such as Oregon and California, the largest component of water storage is snow, and therefore these areas are vulnerable to variations in snowpack due to climatic or other influences. Accurately calculating the area covered by snow in a region is the first to finding the snow water equivalent contained in the snowpack.
The method of snow detection described here is used to calculate the area of snow cover in the Seymour Watershed in North Vancouver. On the Seymour River is a dam from which drinking water is supplied to a large area of Greater Vancouver, and the reservoir here is dependent on the Seymour basin for its water source.
Watershed Snow Cover Example
First, the Seymour Watershed Basin was found on the Landsat image and a polygon mask was created for the total area as shown in Figure 1. This area was delineated using data from the District of North Vancouver water resources website.
Seymour Watershed Basin
Second, the methods developed in this study were applied to the image as shown in Figure 2. Snow covered area was then calculated using the total number of pixels assigned as snow and the Landsat spatial resolution of 30m.
Snow covered area:
(66,938 pixels x 0.030 km2) = 60.24 km2
Seymour Watershed Basin snow cover
Another use for snow cover detection is to analyze and quantify changes over time in snow extent and total area. One field which is particularly interested in changes of snow cover and extent over time is the study of glaciers, including the Arctic sea ice. Glaciers make up a large portion of fresh water storage and fresh water supply, and it has been observed over the last century that the mass of most glaciers is decreasing. Arctic sea ice is an essential characteristic of the Arctic ocean for environmental, economic and political reasons and research into its present and future conditions is increasing. Another important study for snow cover changes is reseach into the ecological implications of changing start and end dates of snow cover in a single year. Some plants and animals depend on snow as environmental indicators of their yearly life cycles.
One issue with estimating snow cover changes over time is that there is approximately 16 days between Landsat repeat images for a particular location. This becomes more difficult in the winter as there is a higher probability of cloud cover which inhibits the viewing of Earth's surface. However, Landsat 4 and 5 Thematic Mapper have over 30 years of imagery collection and therefore still provide useful temporal comparison.
Snow Cover Change Example
These images provide an example of snow cover variation over time:
October 21, 1986
Snow cover: 787.7 km2
Scene ID: LT50470251986293XXX01
November 8, 1993
Snow cover: 5131.3 km2
Scene ID: LT50470251993312XXX02
October 6, 2010
Snow cover: 771.9 km2
LT50470252010279XXX00
October 25, 2011
Snow cover: 4113.4 km2
Scene ID: LT50470252011298XXX01
North Vancouver
North Vancouver
Whistler
Whistler
Whistler
Whistler
Snow
Snow
