Surface Heat Balance Analysis by using ASTER and Formosat-2 data

Soushi KatoDepartment of Earth Sciences,

Earth Dynamic System Research Center,National Cheng Kung University

2008. 3. 11

Introduction

Estimation of surface heat balance in urban areahelpful to understand the causes of heat island

effect

Using Remote Sensing datauseful to obtain spatial pattern

RS data with high spatial resolution and wide spectral coverage are suitable to heat balance estimation

ASTER and Formosat-2ASTER VNIR Formosat-2

False-color images around NCKU, Taiwan (NIR, Green, Red)

Visible & NIR4 band8 m & 2 m resolution

Visible & NIR3 band15m resolution

SWIR TIR6 band 5 band30 m 90 m

Surface Heat Balance

Surface

H

Anthropogenic heatEnergy consumption

Sensible heatSurface-atmosphere

temperature difference

Ground heatSurface-subsurface

temperature difference

Rn LE

Net radiationShort- and longwave

radiation

Latent heatEvapotranspiration

Rn + A = H + LE + GSensible heat flux increase → Heat island phenomenon

AG

Storage Heat Flux (G)

G = G – A = Rn – H – LE

HRn LEG > 0 ⇒ Heat storage

G < 0 ⇒ Heat discharge

• Combine G and A• Estimated by Rn, H and LE

G A

HRn LE

G

Based on the method for ground measurement (e.g. Oke et al., 1999)

Storage heat flux

Netradiation

Sensibleheat

Latentheat

Anthropogenicheat

Groundheat

Ground heat (G)Anthropogenic heat (A)

Difficult to obtain through a wide area

[W/m2]

Net Radiation (Rn) Estimation

　　 　

s : Surface emissivity (absorptance) (Ogawa et al., 2003)

Ta : Atmospheric temperature (K) 　　Ts : Surface temperature (K)

Relative humidity (%)

Reflectance (Liang, 2000)

Rs : Solar radiation (W/m2)

: Albedo

RL : Longwave radiation (W/m2)

= i Ti4 (Stefan-Boltzmann’s law)

a : Atmospheric emissivity (Prata, 1996)

Atmospheric temperature (K)

Rn = (1 ｰ ) Rs + s RL↓ ｰ RL↑

Rs

RL↓ RL↑

s

Sensible Heat Flux (H) Estimation

: Air density (kg/m3)

Cp : Specific heat of air at constant pressure (J/kg K)

Ts : Surface temperature (K)Ta : Atmospheric temperature (K)

ra : Aerodynamic resistance (s/m)

Bulk resistance approach

Wind speed (m/s)

Roughness length (m) 　 Surface type(Brutsaert, 1982; Kondo, 1994; Yasuda, 1995)

Ta 　　　

Ts 　　　　　　 ra 　　　

Ts – Ta

raH = Cp

Latent Heat Flux (LE) Estimation

: Air density (kg/m3)

Cp : Specific heat of air at constant pressure (J/kg K)　

es* : Saturation vapor

pressure (hPa)

ea : Vapor pressure (hPa)

: Psychrometric constant (hPa/k)

ra : Aerodynamic resisntance (s/m)

rs : Stomatal resistance (s/m) 　 (Nishida et al., 2003)

Air temperature (K)

Relative humidity (%)

Surface temperature (K)

ea

es*

ra　

rs

Bulk resistance approachLE =

Cp

es

* – ea

ra + rs

Air temperature (K)

Solar radiation (W/m2)

Minimum rs (s/m)

　　 Surface type

Data Usedfalse color (NIR: Green: Red)

0 5km

Meteorological Station

Study area Tainan, Taiwan

Satellite dataASTER 2000 / 3 / 6 Formosat-2 2004 / 7 / 12

Meteorological data

Tainan 2000 / 3 / 6

Surface Classification Map

Urban

Road

Water

Bare soil

Short grass

Tall grass

Bush

Forest

Formosat-2

Derived from Maximum likelihood method and manual correction

Comparison of Classification Maps

Formosat-2

Roads and vegetations are distinguished more clearly

ASTERSurface classification map around Tainan Station

0 1km

40 　　　 360 W/m2

0 　　　 90 W/m2

-5 　　　 120 W/m2

Heat Fluxes by ASTER & Formosat-2Sensible heat H Latent heat LE

Storage heat G

Net radiation Rn

190 　　　 360 W/m2

Further Study• Usage of ASTER and Formosat-2 data acquired on the

same (at least closer) dates• 2-m resolution pan-sharpened Formosat-2 image

ASTER 15m Formosat-2 8m Formosat-2 2m

Thank you for your attention.