A « European » Perspective Short term: SMOS L2 processor validation Long term: DA in an hydrological models.

Post on 28-Dec-2015

222 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

Transcript

A « European » A « European » PerspectivePerspective

Short term: SMOS L2 processor Short term: SMOS L2 processor validationvalidation

Long term: DA in an hydrological Long term: DA in an hydrological modelsmodels

SMOS Data typeSMOS Data type

►multi-angular; dual polarization L bandmulti-angular; dual polarization L band► SM 4 % vol; 3 day revisitSM 4 % vol; 3 day revisit► Spatial resolution better than 50 kmSpatial resolution better than 50 km► Launched 09 2007Launched 09 2007► Level 1: brightness temperature at H Level 1: brightness temperature at H

and V polarisationand V polarisation► Level 2: daily soil moisture and ocean Level 2: daily soil moisture and ocean

salinity (swath) maps at basic temporal salinity (swath) maps at basic temporal and spatial resolutionsand spatial resolutions

2 / 1 parameter retrieval2 / 1 parameter retrieval

SMOS L2SMOS L2AlgorithmAlgorithm

SM L2 retrieval: SM L2 retrieval: simplificationssimplifications

►200 LCC (ECOCLIMAP) aggregated into 200 LCC (ECOCLIMAP) aggregated into 1010

►Auxillary data not always availableAuxillary data not always available►No direct model for every LCCNo direct model for every LCC►Lack of info on very dry soils / sandy Lack of info on very dry soils / sandy

soilssoils►Only a limited number of emission Only a limited number of emission

sourcessources> Algorithm Theoretical Basis Document (ATBD)> Algorithm Theoretical Basis Document (ATBD)1st issue 12/2005 last issue 12/20061st issue 12/2005 last issue 12/2006

How NAFE can contribute ?How NAFE can contribute ?

► Scaling: do simple strategies/algorithms Scaling: do simple strategies/algorithms (such as the L2 SM processor) perform well ?(such as the L2 SM processor) perform well ?

► A lot of the RT parameters have been A lot of the RT parameters have been calibrated on local experimentscalibrated on local experiments SMOSREX = « typical » grasslandSMOSREX = « typical » grassland « La Londe » = « typical » forest etc« La Londe » = « typical » forest etc -> need to be checked at larger scales / other -> need to be checked at larger scales / other

locationslocations

► Interested in typical scales o(>1km) with Interested in typical scales o(>1km) with sub-pixel variability of o(<1km)sub-pixel variability of o(<1km)

CESBIO/INRA: CESBIO/INRA: short termshort term► AMSR: radiometric extrapolation C -> L AMSR: radiometric extrapolation C -> L (Philippe (Philippe

Richaume and Richaume and Patricia De RosnayPatricia De Rosnay))► Intercomparison of 2 soil emissivity models for soil Intercomparison of 2 soil emissivity models for soil

moisture retrieval algorithms moisture retrieval algorithms (Patricia De Rosnay)(Patricia De Rosnay) the tau-omega model and a coherent model; the tau-omega model and a coherent model; focus on AMSR data;focus on AMSR data; compare the SM inversion performance with what one compare the SM inversion performance with what one

gets in sahelian environments (African Monsoon project) gets in sahelian environments (African Monsoon project) ► Forward RT modelling for forested envt, Litter Forward RT modelling for forested envt, Litter

effects and litter/surface soil moisture relationship effects and litter/surface soil moisture relationship (Jennifer Grant)(Jennifer Grant)

► L2 algorithm evaluation over several NAFE LCC L2 algorithm evaluation over several NAFE LCC (?)(?)

CESBIO/INRA: CESBIO/INRA: long termlong term

► ““Development and evaluation of Development and evaluation of disagregationdisagregation and and assimilationassimilation methods of methods of SMOS radiobrightness into hydrological SMOS radiobrightness into hydrological models”models”

► Multispectral RS Data AssimilationMultispectral RS Data Assimilation TIR (disagregation and/or data assimilation) TIR (disagregation and/or data assimilation) L-band (SMOS resolution)L-band (SMOS resolution) NDVI (forcing/disagregation)NDVI (forcing/disagregation)

► DA of RS data directly, or disaggregated SM ?DA of RS data directly, or disaggregated SM ?► Evaluation on variables (e.g. root zone SM) Evaluation on variables (e.g. root zone SM)

and fluxes (e.g. streamflow)and fluxes (e.g. streamflow)

Disagregation studyDisagregation study►What’s written in the 2005 proposals:What’s written in the 2005 proposals:

““1 km airborne brightness temperature maps will 1 km airborne brightness temperature maps will be aggregated to produce a coarse-resolution be aggregated to produce a coarse-resolution brightness temperature corresponding to the size brightness temperature corresponding to the size of one SMOS pixel.of one SMOS pixel.

A geostatistical procedure based on LC, Soil Type A geostatistical procedure based on LC, Soil Type and Topography will be used together with the and Topography will be used together with the SMOS generated brightness temperature to SMOS generated brightness temperature to produce field scale surface soil moisture mapsproduce field scale surface soil moisture maps

Surface soil moisture maps produced will be Surface soil moisture maps produced will be compared to soil moisture fields obtained by compared to soil moisture fields obtained by inversion of the L-band brightness temperature inversion of the L-band brightness temperature maps acquired by the aircraft as well as ground-maps acquired by the aircraft as well as ground-based soil moisture measurements.”based soil moisture measurements.”

(Philippe Maisongrande ?)(Philippe Maisongrande ?)

Data Assimilation: NAFE’05 Data Assimilation: NAFE’05 datadata

► A coupled SVAT-Lateral redistribution model A coupled SVAT-Lateral redistribution model (ISBA3L/TOPMODEL/Isochrones) will be (ISBA3L/TOPMODEL/Isochrones) will be calibrated against streamflow and in-situ soil calibrated against streamflow and in-situ soil moisture data obtained for the network of moisture data obtained for the network of continuous monitoring sites (continuous monitoring sites (Hélène Roux Hélène Roux is is building up on previous work by Jennifer building up on previous work by Jennifer Pellenq + ?)Pellenq + ?)

► The performance of the data assimilation The performance of the data assimilation methods will be assessed by comparing methods will be assessed by comparing predictions and measurements of streamflow predictions and measurements of streamflow and soil moisture content during NAFE’05and soil moisture content during NAFE’05 HR TIR info alone HR TIR info alone (Hélène ?)(Hélène ?) disagregated L-band / SM info alone disagregated L-band / SM info alone (Olivier ?)(Olivier ?) combined HR TIR/ LR L-band info combined HR TIR/ LR L-band info (Gilles ?)(Gilles ?)

Proposals/FundingsProposals/Fundings► Chances to get a coSMOS3 campaign during Chances to get a coSMOS3 campaign during

NAFE’06 are smallNAFE’06 are small► ESA cal/val project: provides ESA data but no $ !ESA cal/val project: provides ESA data but no $ !

Jetse is PI / Jeff and Gilles co-IJetse is PI / Jeff and Gilles co-I Data will be accessible for validation after launchData will be accessible for validation after launch

► CNES TOSCA 2005 application:CNES TOSCA 2005 application: funding for at least 2 pers. for ~3 weeks for NAFE’06funding for at least 2 pers. for ~3 weeks for NAFE’06 Focus on SMOS calib. for TOSCA 2006 applicationFocus on SMOS calib. for TOSCA 2006 application

► SMOS data flow should start by the end of 2007SMOS data flow should start by the end of 2007► Commissioning phase ends up ~6 months after Commissioning phase ends up ~6 months after

launchlaunch

NAFE’06 ?NAFE’06 ?

► if Murrumbidgee wins: it sounds like if Murrumbidgee wins: it sounds like we’ll replace « streamflow » by « we’ll replace « streamflow » by « evaporation » in the proposal, but evaporation » in the proposal, but objectives won’t change much (or I’m objectives won’t change much (or I’m totally wrong ???) > still very relevant totally wrong ???) > still very relevant for SMOSfor SMOS

► if Goulburn wins: cool, we’ll build on if Goulburn wins: cool, we’ll build on previous studies ! (good for long-term previous studies ! (good for long-term DA)DA)

top related