V. B. Sochava Institute of Geography SB RAS, Irkutsk Study area - PowerPoint PPT Presentation

Modeling of Air Temperature at Tunkinskaya Hollow with The Use of Landsat Space Images N.N.Voropay, E.I.Istomina, O.V.Vasilenko V. B. Sochava Institute of Geography SB RAS, Irkutsk

Study area

Characteristics of altitudinal zones Tunkinskaya Hollow 700 – 800 м Meadows, grasslands, and floodplain 800 – 1000 м Dried and true steppe and forest steppe 1000 – 2000 м Mountain light coniferous taiga 2000 – 2400 м Alpine tundra > 2400 м Loaches belt

Temperature sensors – Termochrones (Dallas Semiconductors, USA)

Landscape maps Tunkinskaya Hollow 1. Landscape map of piedmont area (Bilichenko IN) 2. Landscape map of the central part (Atutova JV) 3. Map of Natural landscapes (Atutova JV) 4. Landscapes of southern East Siberia (Mikheev VS) 2 1 ¡ 4 3

Location sensors installed in 2009 at the Tunkinskaya Hollow

Sensors location

Calculation of temperature from Landsat image T s K 2 / ln(( K 1 / L ) 1 ) = + λ L max L min − L (Qcal - Qcal min) L min λ λ = + λ λ Qcal max - Qcal min Ts – temperature in Kelvin scale, Qcal – calibrated brightness of pixel in 6 th chanel, K1 – constant 1 (equal to 666,09 or 607,76 for ETM+ and TM sensors), K2 – constant 2 (equal to 1282,71 or 1260,56 for ETM+ and TM sensors), Lmax λ и Lmin λ – incoming radiation constant (equal to 12,65 and 3,2 for ETM+ or 15,303 and 1,238 for ТМ sensor), Qcalmin – minimal image digital number (1), Qcalmax – maximal image digital number (255). T v ~kT s T v is air temperature recorded by sensor, n T s - temperature calculated from the satelite image, k - coefficient of proportionality associated with the emissivity of the object and the difference in air temperature and surface.

Calculation of temperature by satellite images (19.07.2010) Tv ~kTs 2) 25,5 25,2 1) 23,7 23,5 3) k= Tv /Ts 22,5 26,0 0,99 17,0 23 1,01 18,0 22,5 1,16 18,0 21,5 0,74 18,6 22 0,80 15,9 22,5 0,84 6-channel of 15,9 20 0,84 Landsat image 14,8 21 0,71 23,7 0,80 24,5 0,71 23,7 23 0,97 23,2 25 1,03 0,93

Calculation of temperature by satellite images 4) k= ax+b, r Correlation values coefficient n Tv ~kTs 5) Calculation of a, b and Tv NDVI -0,58 Tv = (ax+b)Ts , 1 channel 0,37 2 channel 0,63 where Tv is desired value of 3) k= Tv /Ts 3 channel 0,71 air temperature 4 channel -0,15 0,99 5 channel 0,74 1,01 х – brightness in the 7-th 7 channel 0,91 1,16 channel 0,74 0,80 Ts – temperature determined 0,84 from the 6-th channel 0,84 0,71 0,80 0,71 0,97 1,03 0,93

Comparison of temperature, calculated without (left) and with (right), the object emissivity Ts Tv The original image

Thank you for your attention!