To use IJIS

Sitemap

FY2004 Research Products

Investigating the potential of satellite remote sensing data for monitoring the surface temperature regime of permafrost terrain

  • PI: Dr. Vladimir Romanovsky
  • Project title: Investigating the Potential of Satellite Remote Sensing Data for Monitoring the Surface Temperature Regime of Permafrost Terrain

Climate change scenarios predict that global warming will be greatest over high latitudes and that permafrost (or perennially frozen ground - soil frozen for more then two years) areas will be among the regions most heavily affected. One important parameter that controls the ground thermal regime in permafrost areas is the surface heat balance. Satellite remote sensing provides a mean of mapping/monitoring the surface heat balance at the spatial and temporal resolutions required for regional - to continental -scale permafrost investigations. In this project we summarize the results of a comparison between the MODIS land surface temperature (LST) daily global 1km product from Terra and air/near-surface ground temperature measurements from permafrost sites in northern Canada. Results from a tundra site in northern Quebec, Canada (Salluit - 2 stations SILA and TILI fours, and Kangiqsualujjuaq -1 station) show an excellent match between LST values extracted from MODIS pixeis centered on these sites and those measured at the stations for two full annual cycles (2002 and 2003). The site where the SILA station is located is covered with glacial till and moss vegetation. Barren rock is situated. At the SILA and TIKI stations, weather conditions are very windy so there is no snow accumulation during winter in the vicinity of the stations. The best correlation we obtained is between Surface Temperature of MODIS (LST) and Maximum Daily Air Temperature (MDAT), r = 0.97, at SILA and 0.86 - at TIKI. Using Maximum Ground Temperature, the correlation is still excellent at 2 cm (SILA) and 5 cm (TIKI) below the ground surface, with r = 0.96 and 0.90, respectively. Often, ground temperature is not recorded at stations but only air temperature. This is the case at the Kangiqsualujjuaq station where the mean daily air temperature is still well correlated with LST (r = 0.94) while those for SILA and TIKI are r = 0.96 and r = 0.85 respectively. In those cases, the difference is only 1 - 2 °C at the 95% confidence interval.

Currently, we are comparing the MODIS-derived LST products from Terra and Aqua with field measurements from other sites in northern Canada and Alaska representing varying degrees of landscape heterogeneity(e.g. barren, dry/moist/wet/tundra, open forest near tree line). This will allow us to determine how snow, vegetation, and moisture conditions can change the strength of the relationship between the satellite-based measurement (1 km² pixel) and that recorded in the fields.

Figure 1

Figure 1. Meteorological stations used in this study are located in the image covering Canada, with MODIS image tiles overlain. Tile h14v02 covers two meteorological stations, SILA and TIKI, near the village of Salluit. Tile h14v03 covers the meteorological station near the village of Kangiqsualujjuaq (copyright, V. Romanovsky and S. Hachem).

 Figure 2

Figure 2. Close-up view of Salluit and Kangiqsualujjuaq (copyright, V. Romanovsky and S. Hachem).

 

Figure 3

Figure 3
  1. Relation between maximum air temperatures measured at SILA and land surface temperatures retrived from Terra-MODIS (in Celsius).
  2. Relation between maximum air temperatures measured at TIKI tour and land surface temperatures retrieved from Terra-MODIS (in Celsius).
  3. Relation between mean air temperatures measured at Kangiqsualuijuaq and land surface temperatures retrieved from Terra-MODIS (Celsius).
  4. Relation between maximum ground temperatures measured at SILA tour and land surface temperatures retrived from Terra-MODIS (in Celsius).
  5. Relation between maxmum ground temperatures measured at TIKI tour and land surface temperatures retrieve from Terra-MODIS (in Celsius) [copyright, V. Romanovsky and S. Hachem].