EUMETSAT HSAF EUMETSAT SATELLITE APPLICATIONS FACILITY ON SUPPORT

EUMETSAT H-SAF: EUMETSAT SATELLITE APPLICATIONS FACILITY ON SUPPORT TO OPERATIONAL HYDROLOGY AND WATER MANAGEMENT Paolo Rosci ITAF Met Service H-SAF CDOP Science Manager COSMO GM, 07/09/2011

COSMO General Meeting Rome, 5 -9 September 2011 H-SAF soil moisture products based on METOP-ASCAT scatterometer data H-SAF Soil Moisture Cluster Stefan Hasenauer(1) (prepared), Wolfgang Wagner(1), Barbara Zeiner(2), Alexander Jann(2), Patricia de Rosnay(3), Clement Albergel(3) (1) Institute of Photogrammetry and Remote Sensing (TU Wien), Vienna, Austria (2) Central Institute for Meteorology and Geodynamics (ZAMG), Vienna, Austria (3) European Centre for Medium-Range Weather Forecasts (ECMWF), Reading, UK sh@ipf. tuwien. ac. at www. ipf. tuwien. ac. at/radar

Contents § Satellite Application Facility (SAF) in a nutshell § H-SAF overview § Requirements on satellite soil moisture § Product characteristics and generation chains • Large-scale soil moisture • Small-scale soil moisture • Assimilated soil moisture § Recent validation activities and results • In-situ measurements • Hydrological model data § Conclusions COSMO GM, 7 Sep 2011 H-SAF Soil Moisture 3

EUMETSAT SAF Overview § 8 SAF: Satellite Application Facility § Specialization on topics and themes § Providing products and services to users § Part of the EUMETSAT application ground segment § Location: at the Weather Services in EUMETSAT Member and Co-operating States COSMO GM, 7 Sep 2011 4

EUMETSAT SAF Overview SAF Project 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 SAF in support to Nowcasting & Very Short Range Forecasting SAF on Ocean and Sea Ice SAF on Ozone and Atmospheric Chemistry Monitoring SAF on Numerical Weather Prediction SAF on Climate Monitoring SAF on GRAS Meteorology SAF on Land Surface Analysis H-SAF Development Phase COSMO GM, 7 Sep 2011 Initial Operations Phase SAF Continuous Development and Operations Phase (2007 -2012) H-SAF Soil Moisture 5

The H-SAF Consortium and Partners USAM-CNMCA PROTEZIONE CIVILE CNR -ISAC ITALY TELESPAZIO ZENTRAL ANSTALT FUR METEOROLOGIE UN GEODYNAMIC TU WIEN AUSTRIA EUROPEAN CENTRE FOR MEDIUM-RANGE WEATHER FORECASTS ECMWF FINLAND POLAND TURKEY ILMATIETEEN LAITOS INSTITUTE OF METEOROLOGY AND WATER MANAGEMENT TURKISH STATE METEOROLOGICAL SERVICE COSMO GM, 07/09/2011 BELGIUM ROYAL METEOROLOGICAL INSTITUTE GERMANY BUNDESANSTALT FUR GEWASSERKUNDE FRANCE METEO-FRANCE BULGARIA NATIONAL INSTITUTE OF METEOROLOGY AND HYDROLOGY HUNGARIAN METEO SERVICE SLOVAKIA HYDRO-METEOROLOGICAL INSTITUTE H-SAF Consortium: Lead Entity: Italian Meteorological Service 8 EUMETSAT member states and 3 cooperating 6 States + ECMWF

H-SAF Overall Objectives § to guarantee operational provision of high quality level 2/3 satellitederived products and services for: • PRECIPITATION • SOIL MOISTURE • SNOW to perform independent product validation and assessment of the usefulness of the consolidated/improved products § to put in operations new satellite-derived products § COSMO GM, 07/09/2011 7

The Precipitation task To generate and guarantee operational provision of precipitation measurements with unprecedented quality, thanks to the operational availability of microwave radiometers (SSM/I SSMIS on DMSP), and IR radiometry (SEVIRI) MW temperature and humidity sounders (AMSU-A and AMSU-B or MHS) on NOAA and Met. Op satellites also contribute on providing information on precipitation. COSMO GM, 07/09/2011 8

Precipitation Products Identifier Product Description Status H 01 PR-OBS-1 Precipitation rate at ground by MW conical scanners In development H 02 PR-OBS-2 Precipitation rate at ground by MW cross-track scanners In development H 03 PR-OBS-3 Precipitation rate at ground by GEO/IR supported by LEO/MW Pre-Operational H 04 PR-OBS-4 Precipitation rate at ground by LEO/MW supported by GEO/IR (with flag for phase) In development H 05 PR-OBS-5 Accumulated precipitation at ground by blended MW and IR In development Instantaneous and accumulated precipitation at ground by NWP model Operational Blended SEVIRI Convection area / LEO MW precipitation - NEW PRODUCT In development H 06 PR-ASS-1 H 15 PR-OBS-6 COSMO GM, 07/09/2011 9

The Snow task To generate and guarantee operational provision of snow cover information (detection, status, fractional cover, water equivalent) in near real-time through VIS/IR radiometry and MW radiometry The measurements generate intermediate products on flat and forested areas and mountainous areas in separate operational chains, which are then merged to the final products COSMO GM, 07/09/2011 10

Snow Products Identifier Product Description Status H 10 SN-OBS-1 Snow detection (snow mask) by VIS/IR radiometry In development H 11 SN-OBS-2 Snow status (dry/wet) by MW radiometry In development H 12 SN-OBS-3 Effective snow cover by VIS/IR radiometry In development H 13 SN-OBS-4 Snow water equivalent by MW radiometry In development COSMO GM, 07/09/2011 11

The Soil Moisture task To generate and guarantee operational provision of soil moisture measurements both on the surface layer and in the root region by mean of microwave technique, and in particular the ERS and Met. Op scatterometers, (in the low frequency region (1 -10 GHz)) thanks to the high sensitivity of microwaves to the water content in the soil surface layer The primary instrument for all soil moisture products is ASCAT on Met. Op. COSMO GM, 07/09/2011 12

H-SAF SM User Requirements § Spatio-temporal resolution • Near real-time, • Europe § Specification • Standardized file formats: BUFR, GRIB Resolution Dx Cycle Dt Timeliness d (RMS) (km) (h) SM-OBS-3 (large-scale) SM-OBS-2 (small-scale) 0. 1 3 m m -3 3 -3 m m (assimilated) - 0. 05 0. 03 25 sampling: 12. 5 km 0. 1 1 0. 05 0. 03 depending on vegetation - to be assessed 1 25 1 sampling: 0. 5 km - 80 50 sampling: ~ 12. 5 Km 48 36 24 3 object breakt 2 0. 5 one satellite (Metop) depending on access 48 3 36 24 one satellite (Metop) 30 thresh object breakt thresh 25 depending on vegetation SM-ASS-2 object Unit breakt Product thresh § Application and users • (Operational) Hydrology, • Climate monitoring Accuracy 24 24 range: 3 m below surface 2 0. 5 depending on access 24 24 model output End-user requirements for soil moisture products § Characterisation • Validation programmes, • EUMETSAT reviews COSMO GM 7 Sep 2011 Test sites hydrological validation 13

Basic retrieval assumptions § Information from multi-incidence angle observations from scatterometers Steppe Temperate Zone Mongolia - Choybalsan (115 E 48 N) Russia - Roslav (33 E 54 N) pot. Biomass: 300 g/m 2 pot. Biomass: 900 g/m 2 Tropical Forest CAR - Berberati (16 E 4 N) pot. Biomass: 1920 g/m 2 Soil moisture change 28 Nov - 6 Dec 1993 16 Sep - 24 Sep 1995 15 Apr - 26 Apr 1997 Vegetation change

Mean Soil Moisture Seasonal Cycle COSMO GM, 7 Sep 2011 H-SAF Soil Moisture 15

Soil Moisture Products: SM-OBS-3 § Large-scale surface soil moisture observed by radar scatterometer (TU-Wien/ZAMG) • Sensor: METOP-ASCAT • Cycle: 36 hours for full coverage over Europe, available globally over land • Resolution: 25 km, • Timeliness: 130 min • Dissemination: EUMETCast • Format: BUFR 16

SM-OBS-2 § Small-scale surface soil moisture observed by radar scatterometer (TU-Wien/ZAMG) • Sensor: METOP-ASCAT, together with ENVISAT ASAR • Cycle: 36 hours for full coverage over Europe • Resolution: Basic 25 km, sampling to 1 km • Timeliness: ~130 min • Format: BUFR • Concept: Temporal stability of backscatter with ENVISAT ASAR (C-Band) EGU, 6 April 2011 17

SM-OBS-2 ENVISAT ASAR GM coverage of European Database (March 2011) Left: SM-OBS-1 (25 km ASCAT), right: SM-OBS-2 (1 km downscaled surface soil moisture). No-data values are masked and given a quality flag information. Applying linear regression model: DSSM(local) = c + d * SSM(regional), where: DSSM…downscaled surface soil moisture; c, d…intercept, slope; SSM……surface soil moisture resampled. Correlation between local and regional ENVISAT ASAR GM backscatter 18

ASCAT Root Zone Soil Moisture Retrieval Input: ASCAT SM-OBS: Surface Soil Moisture Index Top soil moisture measured: 0 -2 cm ASCAT SM-DAS-2: Root Zone SM Profile Soil Moisture Vertical Profile Variable of interest for Soil-Plant-Atm interaction, Climate, NWP, hydrological applications Root Zone SM Profile: Accurate retrieval requires to account for physical processes d 1 0 d 2 d 3 d 4 Soil Moist m 3 m-3 x 100 SM-DAS-2 Root zone retrieval based on Data Assimilation

ASCAT SM data assimilation SM-OBS (surface swath) SM-DAS (Root Zone profile global) ECMWF Atmospheric conditions SYNOP T 2 m RH 2 m SM-OBS: ASCAT Surface SM EKF Soil Moisture Analysis SM-DAS-2: Soil Moisture Profile H-SAF CDOP: SM-DAS-2 Production chain Quality Control use data when: - Incidence angle: cell number € [11, 26] - Topographic complexity ≤ 20 - Wetland Fraction ≤ 15 - Noise level ≤ 8 For Surface conditions: - Surface is snow free - Soil not frozen 4 layers: 0 -7 cm 7 -28 cm 28 -100 cm 100 -289 cm

SM-DAS-2 § Volumetric soil moisture (roots region) by assimilation in NWP model (ECMWF) • Sensor: METOP-ASCAT • Format: GRIB (regular reduced Gaussian grid or latitude/longitude grid) • Content: 4 different soil layers (covering the root zone from the surface to 2. 9 m) • Cycle: 24 hours SM-OBS-1 Satellite • Resolution: 16 km (T 1279) observations • Timeliness: 36 hours Integrated Forecast System Conventional observations Surface Data Assimilation System SM-ASS-1 - Soil moisture from ECMWF assimilation ZAMG product data base Fig. 12 - Soilmoisture assimilation productgeneration chain. EGU, 6 April 2011 21

Validation Approaches § Ground based systems (in-situ) • Single campaign datasets • Continuous monitoring sites § e. g. ISMN (International Soil Moisture Network) § Modelled data • Hydrological model data • Soil-vegetation-atmosphere-transfer model § Inter-comparison with other satellite products • Active vs. passive microwave products COSMO GM, 7 Sep 2011 H-SAF Soil Moisture 22

SM-OBS-3 vs. in-situ and model § Continuous rainfall-runoff model (MISDc) § Upper Tiber River, Italy § Assimilation improves runoff prediction Brocca, L. , F. Melone, et al. (2010): "Improving runoff prediction through the assimilation of the ASCAT soil moisture product" Hydrol. Earth Syst. Sci. (14), 4113 -4144. doi: 10. 5194/hessd-7 -4113 -2010 Results with and without ASCAT SWI* assimilation for the NIC (a, b) catchment in the period Jan-2007 – Jun-2009: a) observed rainfall and simulated saturation degree; b) observed versus simulated discharge for the sequence of the most significant flood events occurred in the period. H-SAF Soil Moisture 23

SM-OBS-3 vs. SIM Maps of the mean (left) and standard deviation (right) of the near-surface soil moisture (m 3 m-3) from January 2007 to May 2010, from SIM (left) and ASCAT (right). Draper, C. , Mahfouf, J. -F. , Calvet, J. -C. , and Martin, E. (2010): Assimilation of ASCAT soil moisture products into the SIM hydrological platform. Final Report for the H-SAF Associated Scientist Program (AS 09_01), 50 p. H-SAF Soil Moisture 24

SM-DAS-2 vs. in-situ SM-DAS-2 (0 -7 cm) vs in situ (5 cm) R Bias RMSE ) (-) SBR 0. 81 0. 192 0. 216 URG 0. 88 -0. 100 0. 239 CRD 0. 82 0. 468 0. 474 PRG 0. 86 0. 079 0. 128 CDM 0. 85 -0. 020 0. 096 LHS 0. 82 0. 002 0. 114 SVN 0. 85 0. 103 0. 174 MNT 0. 85 -0. 103 0. 177 SFL 0. 89 0. 111 0. 142 MTM 0. 89 0. 156 0. 180 LZC 0. 79 0. 232 0. 261 NBN 0. 90 0. 113 0. 140 Courtesy of C. Albergel (ECMWF) (- Average : R=0. 85, bias =0. 104, RMSE=0. 195 Relative error = 0. 046 m 3 m-3

Validation compliance § SM-OBS-3 (large-scale) § SM-OBS-2 (small-scale) § SM-DAS-1 (assimilated, now SM-DAS-2) COSMO GM, 7 Sep 2011 H-SAF Soil Moisture 26

The Products documentation Each product is furnished with: PRD: Product requirements document ATDD: Algorithm Theoretical Definition Documentation PUM: Product User Manual PVR: Product Validation Report HVR: Hydrological Validation Report A common validation methodology document is also available COSMO GM - Rome, 07/09/2011 27

Conclusions § H-SAF: new satellite soil moisture products for Europe • Large-scale soil moisture (surface, relative index, BUFR): 01 May 2009 – now • Small scale soil moisture (with ENVISAT ASAR, BUFR): 22 Sep 2009 – now • Assimilated soil moisture (root zone, GRIB): 01 Jul 2008 – 31 Aug 2010 § Validated datasets • First results with in-situ and hydrological model data comparison • Further work ongoing with hydrological model data/data assimilation § Challenges and research needs • Investigating spatial variability and error characterisation of scaling approaches § Contact points § H-SAF project web-page: http: //hsaf. meteoam. it § International Soil Moisture Network: www. ipf. tuwien. ac. at/insitu § SM-OBS-03; SM-OBS-02: Stefan Hasenauer: sh@ipf. tuwien. ac. at, www. ipf. tuwien. ac. at/radar § SM-DAS-02: Patricia de Rosnay: Patricia. Rosnay@ecmwf. int COSMO GM, 5 -9 Sep 2011 H-SAF Soil Moisture 28

Thank you for your attention 29

In-situ data source § www. ipf. tuwien. ac. at/insitu • Freely available, online data viewer H-SAF Soil Moisture 30

Required geographic coverage for H-SAF products. - Required H-SAF coverage: 25 -75°N lat, 25°W - 45°E long.

Organizational scheme of H-SAF CDOP COSMO GM, 07/09/2011

European C-Band Scatterometers ERS COSMO GM, 7 Sep 2011 ASCAT H-SAF Soil Moisture 33

METOP ASCAT § Instrument • Advanced Scatterometer § Frequency • 5. 26 GHz (C-Band) § Polarisation • VV § Spatial Resolution • 25/50 km § Temporal Resolution • 8 -15 files/day over Europe, full coverage within 1. 6 days § Data format • BUFR, native EPS COSMO GM, 5 -9 Sep 2011 H-SAF Soil Moisture 34

CDOP – plans for the near future SM-ASS 2: § § Soil Wetness Index in the root zone resulting from assimilation of SM-OBS- in a NWP model. It will be characterized by: • The potential of several observations, co-assimilated with the ASCAT surface soil moisture in the ECMWF Integrated Forecast System (IFS), will be exploited through a multi-variate data assimilation system • Combined assimilation of precipitation/cloud affected radiances over land with ASCAT soil moisture data => a better consistence between soil moisture and precipitation • Extension of the EKF for assimilation of H-SAF snow products => improvement of soil moisture products and better consistency of the hydrological cycle variables

SM-OBS-3 vs. in-situ and model § ASCAT considered reliable, well reproduced the in-situ‚ observed‘ temporal pattern. § SWI correlation coefficients R >0. 90 § RMSE <0. 035 m³/m³ (linear rescaled) Brocca, L. , F. Melone, et al. (2010): "ASCAT soil wetness index validation through in situ and modeled soil moisture data in central Italy“, Remote Sensing of Environment 114(11): 2745 -2755. doi: 10. 1016/j. rse. 2010. 06. 009 Simulated saturation degree for a layer depth of 3 cm versus the three ASCAT saturation degree products, ms, SWI, and SWI* and for the three investigated sites: a) Vallaccia, b) Cerbara, and c) Spoleto H-SAF Soil Moisture 36

SM-OBS-2 vs. model § Filtered and bias-corrected ASCAT data • Bibeschbach-catchment in Luxembourg, 16 probes § ASCAT has potential to help monitoring how river systems approach critical thresholds, operational perspective Matgen P. , S. Heitz, S. Hasenauer, C. Hissler, L. Brocca, L. Hoffmann, W. Wagner, H. Savenije: „On the potential of METOP ASCAT-derived soil wetness indices as a new aperture for hydrological monitoring and prediction: a field evaluation over EGU, 6 April 2011 Luxembourg” (submitted) H-SAF Soil Moisture 37

SM-OBS-2 vs. in-situ Green: Development Phase, Red: Recent H-SAF Soil Moisture 38

SM-DAS-2 Evaluation and comparison of SM-DAS-2, SM-OBS-2, SMOS, for 219 Stations across 4 continents (SCAN, SMOSMANIA, SWATMEX, OZNET, AMMA, REMEDHUS, UMSUOL, UDC-SMOS) SM-DAS-2 SM-OBS-2 SMOS Correlation 0. 7 0. 53 0. 54 Bias (obs-product) -0. 05 -0. 068 0. 12 RMSD 0. 235 0. 255 0. 243 (Albergel et al, in prep 2011)