FR2.L10.2: VALIDATION OF SMOS: SOME FIRST RESULTS

www.monash.edu

Validation of SMOS: Some first results

Jeffrey WalkerDepartment of Civil Engineering

Sandy Peischl, Mahdi Allahmoradi, Christoph Rüdiger, Dongryeol Ryu, Nan Ye, Damian Barrett, Robert Gurney, Yann Kerr, Ed Kim, John Le Marshall

www.monash.edu

www.monash.edu

The Murrumbidgee catchment

www.monash.edu

Catchment characteristics

www.monash.edu

Permanent monitoring stations

www.monash.edu

Ground sampling strategy

www.monash.edu

Farm surface conditions

www.monash.edu

An airborne SMOS/MODIS simulator

6 x Skye VIS/NIR/SWIR Spectrometers6 x Everest Thermal IR’s

L-band RadiometerM

OD

ISS

MO

S

TIR + Spectral

www.monash.edu

AACES field campaigns: supplementary data

www.monash.edu

AACES field campaigns: validation data

www.monash.edu

www.moisturemap.monash.edu.au/AACES

www.monash.edu

www.monash.edu

Rainfall

Canberra

Gundagai

140mm

115mm

Hay

80mm

www.monash.edu

PLMR: 20 January – 20 February 2010

www.monash.edu

Observations

General Overview

dry – little vegetationsome rain & dry down – little vegetation 150+mm rain – forest

(south) and other veg

Murrumbidgee River

Wagga WaggaCanberra

Mt Kosciuszko National Park and State Forest

H-pol

V-pol

www.monash.edu

Research questions

How much of a SMOS pixel needs to be measured to get a reliable brightness temperature average?

How well do the SMOS L1c and L2 brightness temperatures agree with total coverage aircraft data?

How accurate is the SMOS L2 soil moisture product?

How well can we downscale SMOS data?

How well do our LSMs predict soil moisture variability at 1km resolution?

Can SMOS improve LSM prediction of soil moisture by data assimilation?

www.monash.edu

Research questions

How much of a SMOS pixel needs to be measured to get a reliable brightness temperature average?

How well do the SMOS L1c and L2 brightness temperatures agree with total coverage aircraft data?

How accurate is the SMOS L2 soil moisture product?

How well can we downscale SMOS data?

How well do our LSMs predict soil moisture variability at 1km resolution?

Can SMOS improve LSM prediction of soil moisture by data assimilation?

www.monash.edu

Fractional coverage required?

pixbtransb TT ,,ˆˆ

4K

fraction of total pixel area

SMOS 3dB pixel

Transect

10

www.monash.edu

Fractional coverage required?

www.monash.edu

www.monash.edu

Fractional coverage required?

www.monash.edu

SMOS agreement with aircraft data?

H polarisation

www.monash.edu

Level 1c evaluation: patch 9

H Polarisation V Polarisation

www.monash.edu

Level 2 Tb evaluation: patch 3

H Polarisation V Polarisation

www.monash.edu

Conclusions

Field campaigns for the validation of SMOS should typically aim to cover more than 50% of the SMOS pixel

Preliminary results show that SMOS brightness temperatures agree with aircraft data within spec for v-pol but need improvement in h-pol

Preliminary results show that SMOS L2 processing seems ok for the two pixels assessed

We are planning another campaign for Sept; you are welcome to participate

www.monash.edu

Acknowledgments

3 participants each: CESBIO, Vrije Universiteit Amsterdam

2 participants each: Australian Bureau of Meteorology, Institute of Agrophysics Lublin,

1 participant each: Polish Institute of Technology and Life Science, University of Hamburg/Max-Planck-Institute, Technical University of Denmark

DECC, CSIRO (instrument loan and field assistance)

ESA (part financial support of European participants)Yanco Agricultural Institute (access to lab facilities

and accommodation)All farmers allowing access to their properties