SMOS – MIRAS: to provide global & regional measurements of soil moisture, ocean salinity and ice One of the Earth Explorer’s missions (see also lecture D1L1) 1 st space borne global measurement of salinity – currently only complex model output is available Launch scheduled for April to July 2009 Simulated seasonal (winter) sea-surface salinity map. The units are in practical salinity units (psu) Simulated seasonal soil moisture map (winter) of Europe and Africa. The units are 'cubic metre of water per cubic metre of soil'
Measurement of Soil Moisture in the Surface Zone Soil moisture (SM) is a measure of the amount of water within a given volume of soil and is usually expressed as a percentage. Ground measurements Networks (GEWEX) SW, SWIR, …. THIR…. Low frequency microwaves –Active microwaves Vegetation, roughness Revisit Sensitivity –Passive microwaves antenna issue How measured in the surface zone?
SMOS – MIRAS Soil Moisture Measurement Goals Multi-angular Dual polarisation (H and V) 4 % volume 3 day revisit (Vegetation 7 day) Better than 50 km resolution Global products A new technique (2D interferometry) to provide global measurements from space of key variables (SSS and SM) for the first time Pellarin et al. Le Traon et al.
How will the Measurement Goals be Achieved? Constraining models by global soil moisture and ocean salinity observations estimated from dual-pol., multi-angular, L-band brightness temperature measurement acquired by a 2D interferometer. Nadir path Satellite Spacecraft velocity d N Swath 1000 km 30° = 55° Local incidence angle Earth mm Instrument as proposed: 2-D Y-shaped interferometer 4.5m arms 69 LICEF receivers will allow either H &V or full polarisation acquisition Full polarisation. mode is experimental and utility will be reported as part of Cal/Val experiments Use of full pol. mode impacts on data downlink volume (x2 compared to dual pol.)
Measuring Soil Moisture at L-band Negligible atmospheric attenuation (at L-Band 99% atmospheric transmission) Attenuation from vegetation small (for biomass < 5 kg m-2, which is 65% of the Earth’s land surface) Emission from the Earth shows a large contrast between water and land (signal-to-noise ratio from dry to wet soils) due to the large difference between the dielectric constant of water (ca 80) and dry soil (ca 3.5) Emissivity originates from deeper surface soil layer (at L-band ~5 cm) than for shorter wavelengths Jackson and Schmugge, 1989
Aperture Interferometry Angular resolution provided by separated antennas Correlation products s(1)*s(2) Visibility Functions V(D/ ) Inverse F.T. on V T B ( ) Space sampling requirement : every /2 value at least one time ; hence "thinning" possibilities. At L-band classic radiometers require large steerable antennae, MIRAS offers an alternative design that achieves the appropriate resolution through interferometric processing.
SMOS – MIRAS Hexagonal Foot Print From an altitude of 763 km, the antenna will view an area of almost 3000 km in diameter. However, due to the interferometry principle and the Y-shaped antenna, the field of view is limited to a hexagon-like shape about 1000 km across called the 'alias-free zone'. This area corresponds to observations where there is no ambiguity in the phase-difference. Due to orbit and foot print configuration the coverage will be global every 3 days
SMOS – MIRAS Payload Launch: 2009 Mission duration: 3(+2) years Orbit: sun-synchronous, dawn-dusk, 763km, Inclination 98.4 Mass: 683kg (incl. 28kg hydrazine, Platform 317kg – Payload 366kg) Power: 900W (525 W max. payload consumption, 78 Ah Li-ion battery Launcher: Rocket from Plesetsk, Russia Mission Operation: CNES Proteus Control in Toulouse via S-band link in Kiruna Data Acquisition & Processing: X-band downlink in Villafrance (User Service Villafranca /ESA-ESRIN) Launch: 2009 Mission duration: 3(+2) years Orbit: sun-synchronous, dawn-dusk, 763km, Inclination 98.4 Mass: 683kg (incl. 28kg hydrazine, Platform 317kg – Payload 366kg) Power: 900W (525 W max. payload consumption, 78 Ah Li-ion battery Launcher: Rocket from Plesetsk, Russia Mission Operation: CNES Proteus Control in Toulouse via S-band link in Kiruna Data Acquisition & Processing: X-band downlink in Villafrance (User Service Villafranca /ESA-ESRIN)
How Will Soil Moisture be Retrieved? For vegetated surfaces The retrieval of SM requires ancillary data to evaluate the effect caused by the vegetation. The vegetation optical depth can be related to the amount of water in vegetation which can be estimated from indices such as NDVI. BUT dual- polarised, multi-angle L-band data give the opportunity to retrieve both soil moisture and vegetation optical depth. SMOS SM and vegetation characteristics will be produced by an operational SMOS Level 2 retrieval algorithm which is based on an iterative approach, minimizing a cost function computed from the sum of squared weighted differences between measured and modelled brightness temperature (TB) data, for a variety of incidence angles. For bare soil surfaces SM* = a0 + a1 PR +a2 (TBv-TBh) ; SM* retrieved Surface Soil Moisture PR = (Tbv-TBh)/(TBv+TBh) ; Polarization ratio from MIRAS measurements
Sources of Uncertainty in SM measurement Instrumental errors - Radiometric sensitivity, accuracy, calibration stability - Characterisation of system elements incomplete - Interferometric image reconstruction Surface characteristics - Soil surface roughness (causes increase in backscatter) - Soil texture - Land cover & surface heterogeneity - Dew, intercept, snow - Topography - Litter effect - Surface water Radiofrequency Interference Pixel heterogeneity Due to area of footprint, many types of land use can be expected within a resolution cell, for each surface there will be a different model
Cover Dependent Models Relating Emissivity to SM
Accounting for Uncertainties due to Surface Characteristics Pixel heterogeneity - SMOS pixels are highly heterogeneous - Processor distinguishes between three main types of pixels: land, sea/water and mixed - Land surfaces are classified into 12 categories, aggregated from the ECOCLIMAP land cover map (dry sand/desert, bare soils, natural low vegetation, cropland, dense forest, moderately dense forests, snow covered area, marshes, swamps, wetlands, rocky terrain, maintains, ice, urban) ECOCLIMAP Land use map Auxiliary data for static characteristics - Land/sea mask - Water bodies, rivers - Urban areas - Topography: DEM - Soil texture: FAO data set 5’x5’ - Surface roughness Auxilliary data for dynamic characteristics - Land use map (ECOCLIMAP) - Snow cover extent and status (MODIS-MERIS, ECMWF) - Freezing (weather centres) - Land surface temperature (AVHRR, MODIS) - Atmospheric characteristics (weather centres)
How Homogeneous? Forest vs. Non Forest LC fractions
In Conclusion SM retrievals can be attempted in many areas with varying expected accuracy
SMOS Validation and Retrieval Team (SVRT) – Soil Moisture For SM several sites and dedicated teams worldwide Supported by ground and airborne campaigns and Announcement of Opportunity projects in 2005 There will also be SMOS validation for Sea Surface Salinity
SVRT – Airborne System (AMIRAS) AMIRAS on aircraft in SMOS configuration, 24° from nadir Similar configuration but smaller version of MIRAS (three Y-shaped arms) Being flown to provide SMOS simulations and in support of CAL/VAL activities Like MIRAS, this airborne instrument is able to measure in horizontal as well as vertical polarisations in both dual- and full- polarisation modes. Only 4 LICEF like receivers per arm, MIRAS will carry a total of 69 receivers AMIRAS Brightness temperature in the alias-free field-of-view during its maiden flight over Pensaari island in the Lohja lake west of Helsinki. The brightness contrast between water and land is as expected. Credits: UPC (Polytechnic University of Catalonia)
SVRT in China Workshop in 2008 in Beijing to bring SSS and SM teams together In China Dragon 2 project id. 5252 will specifically address SMOS Soil Moisture CAL/VAL –Area centred on Takla Makan Sand Desert. The specified site is selected for its homogeneity, stability, relatively easy to access and significant accuracy from previous passive microwave studies. –Area centred on Northwest of China will look at localisation of algorithms for China –Lead investigators are Dr. Weiguo Zhang (in photo right) & Prof. Yann Kerr (centre of photo left)
SVRT Ice (Antartica) and Land Surfaces (Europe) courtesy Y. Kerr, CESBIO An L-band radiometer operated near Toulouse, France. Temporal measurements are used to determine the effect of vegetation growth and soil moisture on L-band emissivity. Long-term observations of snow and ice-sheet surfaces will be used for external calibration of MIRAS as the ice sheets exhibit stable microwave emission at 1.4 GHz. courtesy PNRA
Retrievals - Forward Modeling by Iterative Minimisation
SMOS Level-1 Products Level 1A: SMOS reformatted and calibrated observation and housekeeping data in engineering units; physically consolidated in pole-to-pole time- based segments; “calibrated visibilities” Level 1B: output of image reconstruction of SMOS observation measurements; consisting of geo-located vectors of Fourier Component of TB on antenna frame Snapshot-wise information Satellite position and attitude TEC magnitude IGRF magnetic vector Sun illumination angle Direct Sun corrected magnitude Snapshot Overall Radiometric Accuracy Geo-located information grid point information: latitude, longitude, altitude Number of measured values in grid point For each value:Flags (Polarization, Sun, position in the Field of View), BT value, radiometric accuracy, Incidence and Azimuth Angle, Snapshot ID, Footprint size Level 1C: Brightness temperatures Swath; Magnitudes are expressed at Top of Atmosphere; Information is geo-located on a Discrete Global Grid (ISEA 4-9); L1c semi- orbit product (pole to pole) split by Land and Sea grid point Two sets of information available: pixel-wise and snapshot-wise L1C Sea product dual polarization snapshot: Brightness temperature @ H-pol (ESA-ESRIN 2007)
Sea Surface Salinity (SSS) computed at each ISEA grid point for a semi orbit (ascending or descending) Three SSS values (3 models, with uncertainties) Pseudo-dielectric constant retrieved Wind Speed and Sea Surface Temperature used in the retrieval Polarized Tb at 42.5º incidence angle at surface and antenna frame Flags and confidence and science descriptors Soil Moisture (SM) computed at each ISEA grid point for a semi orbit (ascending or descending): SM values, optical thickness, physical temperature, Polarized TB (surface and antenna frame at 42.5º), and dielectric constants All quantities have related uncertainties Flags to indicate presence/absence of features/events of interest such as rocks, topography, snow, RFI Descriptors to describe properties such as number of wild views and mean spatial resolution SMOS L2 SM retrieved over North Africa (Cabot et al 2007) SMOS Level-2 Products
Levels 3 & 4 Daily and Averaged Global SM & SSS Maps Level 3 –global, single instrument Level 4 –global root zone SM, multi-instrument retrieval with US missions 2010 on –will be developed and available through French and Spanish national programmes Simulated Seasonal Maps Winter at top Soil Moisture Simulated Sea Surface Salinity
SMOS Ground Segment Acquisition and Products Distribution From Europe CNES – Toulouse Satellite Operations ESAC – Villafranca Data Processing Ground Segment & X-Band Acquisition Station ESRIN – Frascati User Services/ Quality control/ Facilities, Mission management ESTEC – Noordwijk Post Launch Support Office Kiruna Long-term Archive Reprocessing Centre S-Band Acquisition Svalbard NRT Acquisition Station + Science users + Expert Support Laboratories + NRT users
SMOS Data Access SMOS data will be made available through the ESA category-1 procedure, either through dedicated AOs or registration service online ( http://eopi.esa.int ). http://eopi.esa.int SMOS calibration & validation data will be available via the SMOS Cal/Val portal: http://calvalportal.ceos.org/CalValPortal/welcome.do http://calvalportal.ceos.org/CalValPortal/welcome.do Near-real time products will be available either from the WMO- GTS network or the ESAC FTP server Soil Moisture in situ data will be made available via the SMOS Soil Moisture Network Data Hosting under development at the University of Lisbon ESA campaign data will be available via the campaign database http://earth.esa.int/campaigns and via the SMOS CalVal portal http://earth.esa.int/campaigns