PROGRESS IN GEOGRAPHY ›› 2011, Vol. 30 ›› Issue (9): 1118-1124.doi: 10.11820/dlkxjz.2011.09.006

• Original Articles • Previous Articles     Next Articles

Spatial Scale Analysis of the Methods Using U-TAE Algorithm to Extract the Information of Urban Heat Island

LIU Yupeng1, YANG Bo1, CHEN Chong2   

  1. 1. GIS Reasearch Center, College of Resources and Environmental Sciences, Hunan Normal University, Changsha 410081, China;
    2. International Institute for Earth System Science, Nanjing University, Nanjing 210093, China
  • Received:2010-09-01 Revised:2011-01-01 Online:2011-09-25 Published:2011-09-25

Abstract: U-TAE is an algorithm which uses moving windows of different areas to extract the information of UHI and sets each moving window's threshold according to robust estimation method. This algorithm distinguishes non-UHI area and UHI area according to the cumulative times in which pixels have thermal anomalies and determines the intensity of UHI according to the cumulative times with high or low values. It has dynamic threshold and unbiased characteristics. The U-TAE algorithm is used to evaluate its performance in six different scales: 5×5, 11×11, 25×25, 51×51, 101×101 and 201×201. The performance evaluation includes four aspects: area, intensity, space structure of UHI and stability of large-scale UHI. The results show that the best size of the moving window is 11×11. At this size the information of UHI is more complete and effective than at other sizes.

Key words: Changsha, remote sensing, spatial scale, urban heat island, U-TAE