360 likes | 533 Views
A new method for the validation of satellite soil moisture products through rainfall observations. Luca Brocca 1 , Tommaso Moramarco 1 , Wolfgang Wagner 2 , Wouter Dorigo 2 , Clement Albergel 3 , Simone Gabellani 4.
E N D
A new method for the validation of satellite soil moisture products through rainfall observations Luca Brocca1,Tommaso Moramarco1,Wolfgang Wagner2, Wouter Dorigo2, Clement Albergel3, Simone Gabellani4 1 Research Institute for Geo-Hydrological Protection, CNR, Perugia, Italy2 Department of Geodesy and Geoinformation, TU-Wien, Vienna, Austria3ECMWF, Reading, UK4CIMA research foundation, Savona, Italy http://hydrology.irpi.cnr.it Date : 2013/03/20 luca.brocca@irpi.cnr.it
DOING HYDROLOGY BACKWARDS RAINFALL SOIL MOISTURE Infiltrationevapotranspiration
Inverting for p(t): + + Assuming: SM2RAIN – “Doing hydrology backward” relative saturation Evapo-transpiration precipitation runoff drainage Soil water balance equation soil depth Brocca et al., 2013 (GRL)
SM2RAIN MODEL TESTING
SOIL WATER BALANCE MODEL e(t):evapotranspiration s(t):saturationexcess f(t):infiltration Wmax W(t) g(t):percolation SOIL WATER BALANCE MODEL VAL D’AOSTA CENTRAL ITALY Brocca et al., 2013 (HYP), 2013 (VZJ)
HOURLY SYNTHETIC DATA 1-hour 1-day
DAILY SYNTHETIC DATA 1-day 5-day
Italy NS=0.82 Spain NS=0.89 Estimation of daily rainfall for 1-year data France NS=0.81 IN SITU OBSERVATIONS Three sites in Italy, Spain and France with hourly rainfall and soil moisture observations are selected
Italy NS=0.57 Spain NS=0.62 SATELLITE DATA – ASCAT (4-years) Estimation of 4-day rainfall for 4-year data
SM2RAIN VALIDATION OF SATELLITE SOIL MOISTURE PRODUCTS
RAINFALL OBSERVATIONS (ITALY) Precipitation input interpolated by GRISO model: Spatial resolution 2 km temporal resolution 1 hour Raingauges: ~ 3000 Temporal step: 5 - 10 min GRISO Interpolation 1) Maintains the observed punctual rain value on the raingauge spatial position in the interpolated field. 2) Maintains the mean value of the punctual rain observations as the mean value of the interpolated field. Courtesy by CIMA (Gabellani Simone, …) HOURLY OBSERVATIONS FOR THE PERIOD 2010-2011 ASCAT GRID
SATELLITE AND MODELLED SOIL MOISTURE DATA • ASCAT TU-Wien (FTP) • AMSR-E LPRM - asc, desc, asc+desc (VUA) • ESA – CCI SM product • ERA-Land (ECMWF) • and • TRMM 3B42v7 (standard satellite rainfall product) ASCAT GRID ~ 12.5 km
CORRELATION MAPS: ASCAT SV AMSR-E ASCAT 1-day 3-day 5-day AMSR-E descending 1-day 3-day 5-day
CORRELATION MAPS: ASCAT SV AMSR-E ASCAT 1-day 3-day 5-day AMSR-E ascending + descending 1-day 3-day 5-day
CORRELATION MAPS: ASCAT SV AMSR-E 5-day 5-day ASCAT 2010-2011/10/03 ASCAT 2010-2011 5-day AMSR-E
ASCAT vs AMSR-E vs ESA-CCI 1 January 2010 – 31 December 2010 (3-day cumulated rainfall) ESA-CCI ASCAT AMSR-E merged
ASCAT vs ERA-Land vs TRMM 1 January 2010 – 31 December 2010 (3-day cumulated rainfall) TRMM ASCAT ERA Land ASCAT vs ERA
ASCAT vs AMSR-E vs ERA-Land vs ESA-CCI 1 January 2010 – 31 December 2010 (5-day cumulated rainfall) PINK: ASCAT win! CYAN: ASCAT lost!
ASCAT: Pobs vs PTRMM 1 January 2010 – 31 December 2010 (5-day cumulated rainfall) ASCAT Pobs ASCAT PTRMM
ASCAT NOISE vs SM2RAIN 1 January 2010 – 31 December 2010 (3-day cumulated rainfall) ASCAT
SM2RAIN vs RCM SM2RAIN 2010-2011/10/03 TCM 2007-2008 - RMSE ASCAT AMSR-E
SPATIAL CORRELATION ASCAT AMSR-E ascending + descending
TIME SERIES (best pixel) 5-day cumulated rainfall ASCAT AMSR-E descending
TIME SERIES (best pixel) 5-day cumulated rainfall ASCAT AMSR-E ascending + descending
TIME SERIES (central ITALY) 5-day cumulated rainfall ASCAT AMSR-E ascending + descending
TIME SERIES (south ITALY) 5-day cumulated rainfall ASCAT AMSR-E ascending + descending
ESA-CCI vs MODELLED SM (central Italy) Brocca et al., 2013 (JoH, under review)
ESA-CCI vs MODELLED SM (Morocco) Tramblay et al., 2012 (HESS)
r = 0.54 N = 25 r = 0.63 N = 31 r = 0.69 N = 19 ANTECEDENT WETNESS CONDITIONS SICILY – EROSION MODELLING r (in situ vs WACMOS) = 0.81 r (in situ vs ASCAT) = 0.91
CONCLUSIONS The SM2RAIN method can be effectively use to estimate rainfall from soil moisture observations The large-scale and long-term validation of satellite soil moisture products can be carried out by using the SM2RAIN method
Questions? References cited Brocca, L., Melone, F., Moramarco, T., Wagner, W. (2013). A new method for rainfall estimation through soil moisture observations. Geophysical Research Letters, 40, doi:10.1002/grl.50173. Brocca, L., Camici, S., Melone, F., Moramarco, T., Martinez-Fernandez, J., Didon-Lescot, J.-F., Morbidelli, R. (2013). Improving the representation of soil moisture by using a semi-analytical infiltration model. Hydrological Processes, in press, doi:10.1002/hyp.9766. Brocca, L., Tarpanelli, A., Melone, F., Moramarco, T., Caudaro, M., Ratto, S., Ferraris, S., Berni, N., Ponziani, F., Wagner, W., Melzer, T. (2013). Soil moisture estimation in alpine catchments through modelling and satellite observations. Vadose Zone Journal, in press, doi:10.2136/vzj2012.0102. Brocca, L., Zucco, G., Moramarco, T., Morbidelli, R. (...). Modelling soil moisture spatial-temporal variability at catchment scale. submitted to Journal of Hydrology. Tramblay, Y., Bouaicha, R., Brocca, L., Dorigo, W., Bouvier, C., Camici, S., Servat, E. (2012). Estimation of antecedent wetness conditions for flood modelling in Northern Morocco. Hydrology and Earth System Sciences, 16, 4375-4386, doi:10.5194/hess-16-4375-2012. This presentation is available for download at: http://hydrology.irpi.cnr.it/repository/public/presentations/2013/cci-2013-l.-brocca FOR FURTHER INFORMATIONURL: http://hydrology.irpi.cnr.it/people/l.broccaURL IRPI: http://hydrology.irpi.cnr.it Thanks for your attention