Soil Moisture Measurements from Space - · PDF fileSoil Moisture Measurements from Space and...
Transcript of Soil Moisture Measurements from Space - · PDF fileSoil Moisture Measurements from Space and...
Soil Moisture Measurements from Space and Application to Operational Hydrology
Wolfgang [email protected]
Department of Geodesy and Geoinformation (GEO)Vienna University of Technology (TU Wien)
www.ipf.tuwien.ac.at
Soil Moisture
Definition, e.g.
Average
Thin, remotely sensed soil layer
Root zone: layer of interest for most applications
Soil profile
)(m Volume Total)(m VolumeWater
3
3=θ Air
Water
Solid Particles
Cross-section of a soil
∫ ∫⋅
=Area Depth
dzdxdyzyxDepthArea
),,(1 θθ
Microwaves
Microwaves (1 mm – 1 m wavelength)• All-weather, day-round measurement capability• Very sensitive to soil water content below relaxation frequency of water (< 10
GHz)• Penetrate vegetation and soil to some extent
– Penetration depth increases with wavelength
Dielectric constant of water
The dipole moment of water moleculescauses “orientational polarisation”, i.e.a high dielectric constant
Measurement Principles
Radars measure the energy scattered back from the surfaceRadiometers measure the self-emission of the Earth’s surface
SAR
SCAT
ERS-1/2
SAR und scatterometer on European RemoteSensing Satellites ERS-1 and ERS-2
Active Sensors Passive Sensor
5
Microwave missions for soil moisture
33 years of passive and active satellite microwave observations for soil moisture
News Sensors & Products
Recent Successes & Prospects
Recent successes• 2002: First global soil moisture data set from ERS SCAT published• 2003: NASA soil moisture product based on AMSR-E put into operations• 2007: First merged multi-radiometer soil moisture product • 2009: ASCAT soil moisture product available in NRT• 2010: First SMOS soil moisture data released• 2011: International Soil Moisture Network (ISMN) takes off• 2012: First ECV soil moisture data set covering 1978-2010 released
Prospects• 2013: Launch of Sentinel-1
– First operational soil moisture product at ≤ 1 km spatial resolution• 2014: Lauch of SMAP
– First active/passive sensor at L-band
Operational EUMETSAT/H-SAF ASCAT Services
European C-Band Scatterometers
ERS Scatterometer• λ = 5.7 cm• VV Polarization• Resolution: 50 / 25 km
Data availability• ERS-1: 1991-2000• ERS-2: since 1995
–gaps due to loss of gyros (2001) and on-board tape recorder (2003)
• Operations conflict with ERS SAR
METOP Advanced Scatterometer• λ = 5.7 cm• VV Polarization• Resolution: 50 / 25 km
Data availability• >15 years• METOP-A: since 2006• METOP-B: launch in 2012
Daily global scatterometer coverage: ERS (left) and METOP (right)
Backscatter from Vegetated Surfaces
Except for dense forest canopies, backscatter from vegetation is due to surface-, volume- and multiple scattering
Surface scattering(attenuated by
vegetation canopy)
Volume scattering Surface-volume interaction
0000ninteractiosurfacevolumetotal σσσσ ++=
Backscatter versus Soil Moisture & Vegetation
Dense Forest
Grassland& agriculture Grassland &
agriculturewith 30 %
forest cover
Grassland &agriculturewith 60 %
forest cover
Simulations performed with a radiative transfer mixing model
SCAT Seasonal Soil Moisture Dynamics
Mean ERS scatterometer surface soil moisture (1991-2007)
SCAT Noise from Error Propagation
Validation
Available independent data• In-situ measurements• Modelled soil moisture data• Other satellite products
Best Practices• Assessment of absolute product accuracy (RMSE)• Assessment of relative accuracy and anomalies (R, unbiased RMSE)
Methods• Direct comparisons• Triple collocation• R-metric
International Soil Moisture Network (ISMN)
http://www.ipf.tuwien.ac.at/insitu
Soil Moisture Scaling Properties
High variability in time• Remotely sensed layer exposed to atmosphere
Distinct but temporally stable spatial patterns
Temporal stability means that spatial patterns persist in time• Vachaud et al. (1985)
– Practical means of reducing an in-situ soil moisture network to few representative sites
• Vinnikov and Robock (1996)– Large-scale atmosphere-driven soil moisture field– Small-scale land-surface soil moisture field
In-Situ Soil Moisture Time Series
Mean (red) and station (black) in-situ soil moisture time series. REMEDHUS network in Spain. © University of Salamanca
Temporal Stability
( ) ( ) ( ) ( ) ( )tyxyxdyxcydxdtyxA
t prprppr
r ,,,,,,1 θθθ +=′′′′= ∫∫Ρ
Regional scalesoil moisture
Local scalesoil moisture
Linear scaling coefficients
Model Error ≅ 5 %
Scaling by CDF Matching
Mainly due to uncertain soil hydrologic properties and soil moisture scaling properties (temporal stability), absolute soil moisture values from any source (in-situ, model, remote sensing) are highly uncertainBiases are removed by Cumulative Distribution Function (CDF) matching
Source 1 CDF
Source 2 CDF
Scipal, K., Drusch, M., W. Wagner (2008) Assimilation of a ERS scatterometer derived soil moisture index in the ECMWF numerical weather prediction system, Advances in Water Resources, 31, 1101-1112.
ASCAT versus Model
ASCAT versus 3 cm simulated degree of saturation for products, ms, SWI, and SWI* and investigated sites: a) Vallaccia, b) Cerbara, and c) Spoleto.
Brocca, L., Melone, F., Moramarco, T., Wagner, W., & Hasenauer, S. (2010). ASCAT Soil Wetness Index validation through in-situ and modeled soil moisture data in Central Italy. Remote Sensing of Environment, 114, 2745-2755.
ASCAT vs. AMSR-E
Validation over networks in Italy, France, Luxemburg and Spain3 AMSR-E productsASCAT and AMSR-E LPMR performed bestASCAT particularly good in case of anomaly correlations
Brocca, L., Hasenauer, S., Lacava, T., Melone, F., Moramarco, T., Wagner, W., Dorigo, W., Matgen, P., Martínez-Fernández, J., Llorens, P., Latron, J., Martin, C., & Bittelli, M. (2011). Soil moisture estimation through ASCAT and AMSR-E sensors: An intercomparison and validation study accross Europe. Remote Sensing of Environment, 115, 3390-3408
ASCAT
Triple Collocation ASCAT vs. AMSR-E vs. ERA-Interim
New statistical method for the estimation of errors of three data sets• Errors must not be correlated
ASCAT
AMSR-E
GLDAS
Dorigo, W.A., Scipal, K., Parinussa, R.M., Liu, Y.Y., Wagner, W., de Jeu, R.A.M., & Naeimi, V. (2010). Error characterisation of global active and passive microwave soil moisture datasets. Hydrology and Earth System Sciences, 14, 2605-2616
AMSR-E better::ASCAT better
Validation of SMOS and ASCAT over France
Validation over SMOSMANIA networkAn
omal
y
Parrens, M., E. Zakharova, S. Lafont, J.-C. Calvet, Y. Kerr, W. Wagner, J.-P. Wigneron (2012) Comparing soil moisture retrievals from SMOS and ASCAT over France, Hydrology and Earth System Sciences, 16, 423-440.
Validation of SMOS and ASCAT (Operational & Assimilated)
Validation over networks in the US, Europe and Australia• (Initial) operational ASCAT product performs similar to SMOS• Assimilation improves correlation to in-situ data
Correlations between ASCAT and in situ data against correlations between ECMWF SM-DAS-2 and in situ data, then same for SMOS and SM-DAS-2, SMOS and ASCAT.
Same as above but for anomaly correlation values instead of normalised time series.
Albergel, C., de Rosnay, P., Gruhier, C., Muñoz-Sabater, J., Hasenauer, S., Isaksen, L., Kerr, Y., & W., W. (2012). Evaluation of remotely sensed and modelled soil moisture products using global ground-based in situ observations. Remote Sensing of Environment, 118, 215-226
Time Series Correlation ASCAT versus SMOS
Applications
Monitoring of extreme hydrologic eventsRunoff forecastingData assimilationNumerical Weather PredictionLandslide monitoringVegetation monitoringAgricultural monitoringEpidemiological predictionGHG budgetClimate studiesGround water modelling
Monitoring of Extreme Events: Heavy Rain & Hail
Heavy Rain & Hail on 1/7/2012
Monitoring of Extreme Events: Droughts
SCAT Soil Moisture versus River Runoff
Sambesi – Nana‘s Farm 60 days shift
RunoffShifted Runoff
Time
Run
off
So il Mois ture (Blu e L ine)
Scipal, K., C. Scheffler, W. Wagner (2005) Soil moisture-runoff relation at the catchment scale as observed with coarse resolution microwave remote sensing, Hydrology and Earth System Sciences, 9(3), 173-183.
Time Shift andCatchment Size
Event-based Rainfall-Runoff Modelling
"Curve Number Method"
( ) SWIdcSSPSPQ ⋅+=
+−
= with8.0
2.0 2Q … runoffP … precipitationS … retentionSWI … Soil Water Index
Brocca L, Melone F, Moramarco T, et al. (2009) Antecedent wetness conditions based on ERS scatterometer data, Journal of Hydrology, 364(1-2), 73-87.
Improving Runoff Prediction through Assimilation
If improvement can be achieved depends on the quality of the in situ measurements and the applied calibration and data assimilation strategies
Cumulated runoff for the observed and simulated data with and without ASCAT SWI assimilation for four catchments in Italy.
Brocca, L., Melone, F., Moramarco, T., Wagner, W., Naeimi, V., Bartalis, Z., & Hasenauer, S. (2010). Improving runoff prediction through the assimilation of the ASCAT soil moisture product. Hydrology and Earth System Sciences, 14, 1881-1893
Improved Soil Moisture Estimates through Assimilation
Draper, C.S., Reichle, R.H., De Lannoy, G.J.M., & Liu, Q. (2012). Assimilation of passive and active microwave soil moisture retrievals. Geophysical Research Letters, 39, L04401
ASCAT Assimilation at NWP Centres
Impact on temperature and humidityCentres
• ECMWF• Mete Office• Meteo France• ZAMG
AustraliaTropics
North America NorthernHemisphere
Skill of relative humidity forecasts
Dharssi, I., Bovis, K.J., Macpherson, B., & Jones, C.P. (2011). Operational assimilation of ASCAT surface soil wetness at the Met Office. Hydrology and Earth System Sciences, 15, 2729-2746
Afternoon rain more likely over drier soils
Taylor, C.M., de Jeu, R.A.M., Guichard, F., Harris, P.P., & Dorigo, W.A. (2012). Afternoon rain more likely over drier soils. Nature, 489, 423-426.
Dark red = rain falls most likely over drier soils
Pink = rain falls likely over drier soils
Blue = rain falls likely over wetter soils
Dark blue = rain most falls likely over wetter soils
Prediction of Water-Borne Diseases
Unpublished results of Brocca, Montosi & Montanari
Landslide Monitoring
Brocca, L., F. Ponziani, T. Moramarco, F. Melone, N. Berni, W. Wagner (2012) Improving Landslide Forecasting Using ASCAT-Derived Soil Moisture Data: A Case Study of the Torgiovannetto Landslide in Central Italy, Remote Sensing, 4(5), 1232-1244.
Comparison between observed (circles) and estimated (triangles) crack aperture of the Torgiovannetto Landslidein Central Italy from the beginning to the end of the selected rainfall events.
Climate Change: Merging of Multi-Satellite Data Sets
E R S -1/2 S C A T
1991-2011
M E TO P A S C A T
2006-now
EO S /A cquaA M S R -E
2002-now
C oriolisW indsat
2003-now
T R M MTM I
1998-now
D M SPS S M /I
1987-now
N im b us 7S M M R
1978-1987
A ctiveLevel 2
R etrieva l
A ctiveLevel 2
R etrieva l
P ass ive Level 2
R etrieva l
P ass ive Level 2
R etrieva l
P ass ive Level 2
R etrieva l
P ass ive Level 2
R etrieva l
P ass ive Level 2
R etrieva l
A ctive E C V D ata1991-now
Passive E C V D ata1978-now
R eferenceG LD A S,
E R A -In terim
M erg ingA ctive
Level 2 D ata
M erg ingP ass ive Level 2
D ata
M erg ingA ctive-P ass ive
D ata Sets
A ctive-Passive
E C V D ata1978 -now
L evel 1D ata Sets
E CVProduction
System
Ancilla ry D ata
F or Ac tive R etrieva l
Anc illa ry D ata
For P assive Retrieva l
Trend 1988-2010
Dorigo et al. (1988-2010) in harmonized multi-satellite surface soil moisture, Geophysical Research Letters, 39, L18405, 1-7.
Conistent Trends in Modelled Soil Moisture (top), Precipitation (middle) and NDVI (bottom) Time Series
Dorigo et al. (1988-2010) in harmonized multi-satellite surface soil moisture, Geophysical Research Letters, 39, L18405, 1-7.
Outlook
Availability and quality of satellite soil moisture data sets is improving• METOP-A + METOP-B• AMSR-2
Future Opportunities• Sentinel-1
– 1 km soil moisture and freeze/thaw– 30 m water bodies
• Soil Moisture Active Passive (SMAP)
Early Announcement• Satellite Soil Moisture Validation and Application Workshop
– 1-3 July 2013 at ESA ESRIN, Frascati, Italy– Co-organisation ESA, EUMETSAT, WMO, GEWEX, CEOS and GCOS