The EUMETSAT Satellite Application Facility on Land Surface Analysis (Land SAF): Proposed Products

Transcription

1 The EUMETSAT Satellite Application Facility on Land Surface Analysis (Land SAF): Proposed Products by: Alirio Arboleda Based on material provided by Carla Barroso (IPMA)

2 OBJECTIVE Introduce the EUMETSAT s Satellite Application Facility (SAF) on Land Surface Analysis (Land SAF) and present the products developed, their characteristics and some validation results Invite you to use the Land SAF products in your projects and applications and to participate to the improvement of these products by given feedback to products developers

3 EUMETSAT SAF NETWORK Satellite Application Facility (SAF) is an European scientific network designed by EUMETSAT. Objectives Improve EUMETSAT s Member States exploitation of satellite data Encourage the utilisation of existing skills and infrastructure in Member States and Cooperating States Cost-effective exploitation: Services are distributed in the most appropriate way Foster development of cooperation with non-member States and other organisations

4 Objectives SAF on Land Surface Analysis (LSA) Develop techniques that allow an effective use of remote sensing data (MSG, EPS) data related to: LAND LAND-ATMOSPHERE Interactions BIOSPHERIC Applications Timely provide: Products User support Organisation principles Algorithms developed at one of the participating Institutes Algorithms handed over to IM for integration and production LSA-SAF consortium Instituto de Meteorologia (IM), Portugal Meteo-France (MF), France Royal Meteorological Institute (RMI), Belgium [Finnish Meteorological Institute (FMI), Finland] IMK, University of Karlsruhe IDL, University of Lisbon UV, University of Valencia CDOP-2 new members: KCL, UK VITO, Belgium

5 LSA-SAF chronogram of activities Meteosat-8 (MSG-1) MetOp-A Meteosat-9 (MSG-2) MetOp-B Meteosat-10 (MSG-3) MetOp-C Meteosat-11 (MSG-4) MTG-1 MTG-S Dev. Phase: Sep 1999 Initial Operations Phase: Feb 2005 Continuous Development & Operations Phase: Mar 2007 Continuous Development & Operations Phase 2: Mar 2012

6 LSA-SAF MSG Products Surface Radiation Surface Water Balance LST Albedo LongWave Flux ShortWave Flux Snow Cover Evapotranspiration Vegetation Wild fires Fraction Veg Cover LAI FAPAR NDVI from Metop Fire Risk Mapping(Europe) Fire Radiative Power Fire Detection & Monitoring Development Pre. Operat. Operational Increased level of maturity

7 Common Products Characteristics All products have a quality flag and/or error bar associated All products have detailed documentation (Product User Manual, Validation Report, Algorithm theoretical Basis Document) 4 production areas for MSG - Europe - Northern Africa - Southern Africa - Southern America Euro NAfr SEVIRI resolution Variable time resolution -15 min to 10 days SAme SAfr EPS products generation for a subset of variables(lst, DSLF, NDVI)

8 LSA-SAF MSG Products Surface Radiation LST Albedo LongWave Flux (DSLF) ShortWave Flux (DSSF) Development Pre. Operat. Operational Increased level of maturity

9 Radiation Products Surface Radiation Budget F rad = F SW + F LW F F SW SW F SW F F LW F SW LW F rad = DSSF - ALBEDO DSSF + DSLF - DSLF (1-ε)-εσ εσlst 4 Short-wave Radiation budget at the surface Long-wave Radiation budget at the surface The surface radiation budget is the balance between the incoming energy from the sun and the outgoing energy of Earth.

10 Radiation Products - DSSF Downwelling Surface Shortwave Flux Radiative energy flux in the wavelength interval [0.3µm, 4.0µm] reaching the Earth's surface per time and surface unit. LSA SAF method (Geiger et al., 2008) Input data: 0.6, 0.8 and 1.6 µm SEVIRI channels TCWV from ECMWF Cloud mask from NWCSAF

11 Radiation Products - DSSF Methodology DSSF S 0 v( t)cosθ sun T effective Solar Constant Distance sun-earth (jday) Solar zenith angle Effective transmittance T effective = T atm 1 1 A surf Atmosphere transmittance A atm The effective transmittance of the atmosphere is a funtion of atmospheric constituents Cloud transmittance T effective = T atm 1 T T cloud atm _ below_ cloud A surf A cloud Simplified physical description of the radiation transfer in the cloud-atmospheresurface system Surface albedo Cloud albedo

12 Radiation Products - DSSF Characteristics Product Horizont al Resoluti on & Coverag e Temporal Resolution Genera tion Freque ncy Target Accura cy DSSF MSG disk Instantaneo us & Daily 30 min & Daily 5-10 % Instantaneous (30 min) DSSF => completed with METOP/AVHRR to improve high latitudes coverage

13 Radiation Products - DSSF Monthly variation of daily accumulated Solar Radiation 15 Jan March May July Sep Nov 2009

14 Radiation Products - DSSF Validation Validation at measurement sites

15 Radiation Products - DSSF Validation 2004 to 2006 Clear sky: biais : 5 W/m2 stdev : 40 W/m2 Cloudy sky: biais : 5 W/m2 stdev : 115 W/m2

16 Radiation Products - Albedo Albedo (AL): Fraction of incident radiation reflected by the surface High Albedo Values over Snow Albedo Seasonal Cycle: Snow Primarily driven by Vegetation & Snow Deserts Low Albedo Values Vegetated Areas

17 1.6µm Radiation Products - Albedo Methodology (Input data) 0.6µm, 0.8µm, 1.6µm SEVIRI channels Solar and View Angles Land/Sea Mask Cloud Mask (SAF-NWC software) Total Column Water Vapour, and Pressure (ECMWF) Ozone Content (Climatology) Aerosol Optical Thickness (Climatology)

18 Radiation Products - Albedo Methodology TOA-Radiances 0.6µm, 0.8µm, 1.6µm for each day for each slot Geometry Geometry Atmospheric Correction (SMAC) TOC-Reflectances 0.6µm, 0.8µm, 1.6µm BRDF Model Inversion Model Parameters 0.6µm, 0.8µm, 1.6µm Angular Integration Cloud Mask Atmospheric Constituents Previous Model Estimate Spectral Albedo 0.6µm, 0.8µm, 1.6µm Spectral Integration Broadband Albedo [0.3µm,4µm], [0.4µm,0.7µm], [0.7µm,4µm] Lucht and Roujean (2000), Geiger et al. (2008)

19 Radiation Products - Albedo Validation boreal forest Modis SW-BH albedo MSG albedo ground measurements mixed shrub/tree

20 Radiation Products - Albedo Validation MSG albedo ground measurements Modis albedo Aerosol optical thickness Temporal evolution of the albedo estimate is influenced by rainfall. Spurious fluctuations appear to be caused by aerosol effects.

21 Radiation Products - Albedo Validation Comparison in Central Africa (Savannah) Good correlation between LSA SAF and MODIS albedo

22 Radiation Products - LST LST Land Surface Temperature (LST) is the radiative skin temperature over land. Generation Frequency - 15 min clear sky pixels... over land... where estimated errors < 4K Available since 2005

23 Radiation Products - LST Methodology Generalised Split-Windows (Wan and Dozier 1996, adapted to SEVIRI: Trigo et al., 2008a): TOA Clear sky brightness temperature at 10.8µm and 12.0µm (SEVIRI) 1 ε ε T + T 1 ε ε T T LST = (A + A2 + A3 ) + (B1 + B2 + B3 ) 2 2 ε ε 2 ε ε C GSW parameters depend on: 1. total column of water vapour 2. satellite viewing angle ECMWF fc ε = average ε = difference of channel emissivities

24 Radiation Products - LST Methodology Channel Emissivity From fraction of Vegetation Cover Pixel MSG ε = ε veg FVC+ ε ground (1-FVC) LSA SAF Product Emissivity is estimated as a weighted average of that of bare ground and vegetation elements within the pixel

25 Radiation Products - LST Validation at Evora (Pt) Evora Day time (ºC) BIAS RMSD SEVIRI MODIS Night time (ºC) BIAS RMSD SEVIRI MODIS Trigo, I. F., I. T. Monteiro, F. Olesen, and E. Kabsch, (2008) in J. Geophys. Res., 113

26 Radiation Products - LST Validation at Gobabe (NA) May 2008 Jul 2008 Nov Obs In Situ (ºC) Bias: -0.31ºC RMS: 1.16ºC LST SEVIRI/MSG (ºC) Obs In Situ (ºC) Bias: -0.91ºC RMS: 1.67ºC LST SEVIRI/MSG (ºC) Obs In Situ (ºC) Bias: +0.24ºC RMS: 1. 44ºC LST SEVIRI/MSG (ºC) Freitas, S. C., I. F. Trigo, J. M. Bioucas-Dias, F.M. Goettche, 2010 in IEEE TGRS

27 Radiation Products - LST Validation: Comparison between MSG-LST and MODIS LST (Central Africa Savannah) O SEVIRI LST * MODIS LST Daytime Differences depend on viewing angle Directional Character of remotely sensed LST Overall SEVIRI LST is warmer than MODIS LST, with nighttime systematic differences ranging between 0.5-1ºC.

28 Radiation Products - DSLF Downwelling Surface Long-wave Flux (DSLF): Is the thermal irradiance reaching the surface in the thermal infrared spectrum interval [4µm,100µm]. Long-wave Radiation reaching the Surface: Essentially emitted by the lowest 100 m of the atmosphere It is controlled by concentration of absorbing gases CO2, CH4, H2O,... presence of clouds, clouds phase temperature profile & temperature of cloud base...

29 Radiation Products - DSLF Methodology Total irradiance at the surface within the [4 100 µm] interval: Bulk Parameterization based on the Stefan-Boltzman Law: 4 cloudy DSLF = σε skytsky = ndslf + (1 n) DSLF Fractio n of Cloud clear sky w w exp α + β ε T sky = T 2 + ( γ Td 2 + δ ) = m Trigo et al. (2010) Total column water vapour 2m air temperature 2m dew point depression

30 Radiation Products - DSLF Example of instantaneous DSLF product and associated quality flag for 14/03/2014 at 12:00 UTC

31 Radiation Products - DSLF 15 Jan Apr Jul Oct 2011

32 Radiation Products - DSLF Validation: Northern Europe station CERES SEVIRI DSLF clear sky cloudy sky Period Jan Apr 2007 Stations Cambourne, UK Lerwick, UK Toravere, Estonia CERES SEVIRI DSLF Bias RMSE Bias RMSE Clear Sky All Sky Problems: - Cloud identification at the edge of Meteosat disk - DSLF model for ice clouds - Temperature inversions

33 Radiation Products - DSLF Validation: Central Europe station CERES SEVIRI DSLF clear sky cloudy sky Period Jan Apr 2007 Stations Palaiseau, France Payerne, CH Carpentras, France CERES SEVIRI DSLF Bias RMSE Bias RMSE Problems: - temperature inversions Clear Sky All Sky

34 Radiation Products - DSLF Validation: Semi-arid and desert stations CERES SEVIRI DSLF clear sky cloudy sky Period Jan Apr 2007 Stations Tamanrasset, Algeria Sde Boqer, Israel Niamey, Niger CERES SEVIRI DSLF Bias RMSE Bias RMSE Problems: - Impact of high aerosol loads Clear Sky All Sky

35 LSA-SAF MSG Products Surface Water Balance Evapotranspiration Snow Cover Development Pre. Operat. Operational Increased level of maturity

36 Water balance products Evapotranspiration (ET) ET: Flux of water vapour between ground surface and the atmosphere (mm/h) 0 5 (mm) Instantaneous ET, one image every 30 minutes Daily cumulated ET (DMET)

37 Methodology Water balance products Evapotranspiration (ET) A simplified version of ECMWF TESSEL SVAT module, forced with LSA SAF radiation products (DSSF, DSLF & AL) & ECMWF meteorology. MSG Pixel Tuiles: % Bare soil % forests % crops % grass

38 Methodology from satellite MSG-II From database (ECOCLIMAP) AL -Surface ALbedo- DSSF-Shortwave flux at surface DSLF-Longwave flux at surface LAI -Leaf Area Index- FVC -Fractional Vegetation Cover- LST -Land Surface Temperature- T i -Tiles in pixel- FV i -Fraction of tiles- Rs i -Tile minimum stomatal resistance LAI i -Tile Leaf Area Index - FVC i -Tile fractional vegetation cover AL M -Monthly pixel Albedo Water balance products Evapotranspiration (ET) (2) (1) MODEL FORMULATION MET QF From NWP (ECMWF) Ta -Air temperature- U -Wind speed- Pa -Air pressure - Td -Dew-point temperature- ST -Soil Temperature- SM -Soil Moisture- 1 included in last version 2 future versions

39 Validation Water balance products Evapotranspiration (ET) Comparison to observations in different climatic and environmental conditions Comparison to output from other models Good agreement between simulations and observations ; the best agreement is observed in areas dominated by grasslands and mixed forests

40 Water balance products Snow Cover (SC) Different signatures of snow, ice, and clouds on 0.6, 0.8 & 1.6 µm channels reflectances. Non proc Snow Snow-free Unclass Water A thresholding technique is applied to distinguish surfaces covered with snow or ice from clouds and snow-free pixels.

41 LSA-SAF MSG Products Vegetation Fraction Veg Cover LAI FAPAR NDVI from Metop Wild Fires Fire Risk Mapping(Europe) Fire Radiative Power Fire Detection & Monitoring Development Pre. Operat. Operational Increased level of maturity

42 Vegetation Products FVC Fractional Vegetation Cover LAI Leaf Area Index fapar Fraction of Absorbed Photosynthetically Active Radiation Fraction of vegetation on a flat background. Accounts for the surface of leaves contained in a vertical column normalized by its crosssectioanl area. Indicator of the health (productivity) of vegetation.

43 Algorithms FVC is estimated through the application of a spectral mixture analysis methodology (0.6, 0.8 & 1.6 µm channel refletances), developed taking into account the spectral variability of vegetation in different ecosystems (e.g. Bateson et al. 2000; García-Haro et al 2005); LAI is estimated from FVC (Roujean and Lacaze (2002)); FAPAR is based on simulations of surface reflectances in optimal angular geometries (Roujean and Bréon, 1995). BRDF parameters (contain specific spectral directional signatures of vegetation reflectances).

44 Validation results Example of validation exercises carried out in SAfrica: Rapid green-up (Zambia) Minimum biomass Minimum biomass Peak of biomass Peak of biomass

45 NDVI from Metop Near-global, 10-daily composite images (synthesized from the "best available" observations registered in the course of every "dekad" by the orbiting earth observation system Metop-AVHRR)

46 Fire Radiative Power FIRE RADIATIVE POWER (FRP) - The Fire Radiative Power (FRP) is the amount of radiant energy emitted per unit time during a vegetation fire. is related to the rate at which fuel is being consumed

47 Fire Radiative Power By integrating FRP during the lifetime of a vegetation fire we get the total combusted biomass(kg) The FRP product is derived using a Fire Thermal Anomaly (FTA) algorithm. Ii works mainly on statistics derived from the 3.9 µm and 11.0 µm brightness temperatures, and their differences

48 Fire Radiative Power FRP product allows to: Detect an active vegetation fire Compute the radiant energy per unit time for the detected fire Estimate trace gas emissions from the fire

49 Temporal Frequency & Availability Availability Vegetation Products 5-Day 30-Day Daily 10-day >2006 NDVI 10-Day 10-Day >March 2013 ( ) FRP Daily/15 min. Daily/15 min. >2008

50 How to access products files? Publications Download Data Documentation (User Manuals, Validation Reports) Helpdesk

51 How to access products files?

52 How to access products files? 1. register 2. login

53 How to access products files?

54 How to access products files?

55 How to access products files?

56 How to access products files?

57 How to access products files?

58 How to read/visualize the files? Data stored in Hierarchical Data Format 5 (HDF5), a file format designed to store and organize large amounts of numerical data. It is supported by the nonprofit HDF Group The files are compressed with bzip2 ( When downloaded, you should uncompress with bunzip2 Free softwares, eg :HDFview ( Commercial softwares (matlab, idl, )

59 Visualize data with HDFview ( Evapotranspiration values (array) Quality flag (array) This software allows: - Visualize images - Visualize values - Histograms - Save images in other formats (.jpg, bmp, gif,.)

60 Visualize data with HDFview Vizualization with different color palettes Attribute of dataset

61 Read files with matlab Datasets attributes

62 3 datasets for the hdf5 file Daily evapotranspiration Information on the data: attributes

63 Read the HDF5 file (dataset ET) with hdf5read Step 1: use «hdf5read», specifying the dataset to read Step 2: T is an array of integers. You should divide by the scaling factor.

64 Select a point based on its geographical coordinates

65 Select a point based on its geographical coordinates

66 Step 1: choice of the point (Ex.: N, 10 W) Step 3: divide by scale factor Step 2: read file lat-long Step 4: compute the distance between point and pixels Meteosat Chosen point corresponds distance minimum

67 ET in mm/hr Quality flag

68 Extraction of a time series

69 Clear sky day Clouds passing over According to the table in the product user manual (PUM) Nominal quality Below nominal quality Poor quality

70 Daily values

71 Questions/Comments?