Europ

1
A synergetic use of observations from MODIS,
SEVIRI MSG, ASAR and AMSR-E to infer a daily soil
moisture index
C. Notarnicola1, F. Di Giuseppe2, K. Lewinska1,
L. Pasolli1,3, M. Temimi4, B. Ventura1, M.
Zebisch1 1 EURAC-Institute for Applied Remote
Sensing, Viale Druso 1, Bolzano, Italy.
2ARPA-ServizioIdroMeteoClima, Viale
Silvani 6, Bologna, Italy 3Dep. of Information
Engineering and Computer Science, University of
Trento, Via Sommarive, 14, Trento, Italy.
4NOAA-CREST/NOAA-CREST/The City
University of New York, The City College, 140th
St _at_ Convent Ave. Steinman Hall (T-109), New
York, NY 10031.
IEEE International Geoscience and Remote Sensing
Symposium - IGARSS 2011 Vancouver, Canada -
July 24-29, 2011
2
Outline

• Introduction
• Main concept multi-sensor approach
• Test sites and EO data
• Apparent Thermal Inertia (ATI) approach
• Synergy with SAR images and AMSR-E data
• Experimental results
• Conclusions and future steps

3
Introduction

• Techniques and methods to evaluate soil and
  vegetation water content includes as main
  instruments passive and active microwave methods
  but also some indirect measurements based on
  radiometric techniques in the optical-thermal
  range.
• Due to large data set availability from different
  sensors, in the last years the synergy among
  sensors and exploitation of multi-sensor
  approaches has increased notably.
• The objective of this study is to infer a soil
  moisture index (soil moisture classes) from an
  approach mainly based on the concept of apparent
  thermal inertia (ATI) by the following steps
• Exploitation of the Apparent Thermal Inertia
  (ATI) from optical sensors (MODIS)
• Continuous calibration with SAR images (when
  available) and AMSR-E data
• Synergy with SEVIRI MSG acquisitions

4
Main concept multi-sensor approach
5
Test sites and EO data

• Data availability
• MODIS images for year 2007, 2008-2009, 2010
• AMSR-E/Aqua level 3 global daily surface soil
  moisture images
• SAR images (when available on Emilia Romagna test
  sites)
• SEVIRI MSG acquisition contemporary to MODIS
  images

Test sites Dates Product Polarization Swath/? MODIS acquisistion
SCRIVIA Apr. 4, 2008 Oct 31, 2008 IMS IMS VV VV 2/23o 2/23 April 4. 2008 cloudy
MATERA July 13, 2008 Oct.10, 2008 May 7 2008 April 11, 2009 APS APS IMS APS HH/HV HH/HV VV HH/HV 2/23o 2/23 2/23 2/23o July 13, 2008 Oct.10, 2008 May 7, 2008 April11, 2009
CORDEVOLE June 14, 2004 July 19, 2004 Sept. 27, 2004 IMS IMS IMS VV VV VV 2/23o 2/23o 2/23o cloudy July 19, 2004 Sept.27, 2004
6
Apparent Thermal inertia (ATI) algorithm

• Physical Thermal Inertia (TI) (Watson et. al.,
  1971 Price 1977)
• Response to temperature change
• Physical TI v(densitythermal conductivityheat
  capacity)
• Apparent Thermal Inertia (ATI) (Price 1985 Mitra
  Majumdar 2004 Claps Laguardia, 2004)
• ATI (1-albedo) / (Temperature max - Temperature
  min)
• Thermal image pair solar noon and pre-dawn

7
ATI error analysis
If the limit of ??T 10K is considered in order
to have reliable ATI estimates Cai et al.,
2007, the error on ATI is around 0.002 which
corresponds to 5 of the lowest values detected
in this analysis.
8
Cross-comparison with SMC retrieved from SAR
images
ATI-Ground measurements ATI1.810-3SMC0.0154 R2
0.76 (SMC expressed in )
ATI-SMC_SAR ATI1.310-3SMC0.019 R20.78 (SMC
expressed in )
Three classes (Emilia Romagna Test site) ATI lt
0.04, SMC-SAR lt 10 0.04 lt ATI lt 0.05, 10 lt
SMC-SAR lt 15 ATI gt 0.05, SMC-SAR gt 15, OA72
ATI SAR
9
Temporal filtering with AMSR-E data
Even under similar soil moisture conditions and
acquisition time, the ATI values show a high
variability. It is therefore necessary to
introduce a filtering technique to reduce the
noise in the observed data. The use of
microwave based time series of soil moisture to
refine the ATI based product should perform
better than any other stand-alone signal analysis
technique like moving average as the microwave
estimates are intrinsically consistent with ATI
estimates. The main assumption of this study is
the agreement between soil moisture estimates
from microwave and ATI. This expected agreement
fosters using the microwave time series to filer
and refine the ATI product. A temporal moving
window has been considered by using the following
expression

• If there is no significant changes the average
  values of SMCAMSRE(tim) are equal to
  SMCAMSRE(ti), the weight is equal to 1 and the
  filter performs only a simple mean over the days
  considered
• If there is an increase in soil moisture values
  the average values of SMCAMSRE(tim) are higher
  than SMCAMSRE(ti), the variations are enhanced
  and so the effect of averaging of the filter is
  reduced
• If there is a decrease in soil moisture values
  the average values of SMCAMSRE(tim) are lower
  than SMCAMSRE(ti), the variations are reduced and
  then smoothed by the filter, thus reducing the
  noise.

10
Analysis of temporal trends Emilia Romagna
Comparison of the temporal trend among SMC
(cm3/cm3), ATI originally calculated and ATI
filtered. H stands for High NDVI values (gt 0.4)
and L stand for Low NDVI values (lt 0.4).
No filter Simple filter Filter using AMSRE data
NDVI lt 0.4 0.58 0.59 0.72
NDVI gt 0.4 0.45 0.45 0.56
Comparison between ATI and measured soil moisture
values (SMC) over 1 year period for Emilia
Romagna test site. The values represent the
determination coefficients between ATI values and
SMC in the different cases considered.
11
SMC classes from ATI

• From the error analysis on ATI, we need to
  consider
• For the theoretical error the lowest value 0.015
  K-1 has been considered, thus assuming that the
  filtering notably reduces the effect due to
  acquisition time
• The standard deviation values of the different
  ROI considered have a median value around 0.003
  K-1.
• Considering these two independent sources of
  errors, the total error is 0.0153 K-1. As the
  ATI values range from 0.04 K-1 to 0.10/0.12 K-1,
  the total error determines the possibility to
  detect at least four classes.
• In order to determine the class boundaries and to
  verify their consistency with the error on SMC
  measurements (around 5 of the measured value),
  the clustering tool of Maltlab? has been
  considered.

ATI/SMC 1 (lt 0.05) 2 (0.05-0.07) 3 (0.07-0.085) 4 (gt0.085)
1 (lt0.17) 0.57 0.43 - -
2 (0.17-0.25) 0.24 0.62 0.09 0.05
3 (0.25-0.3) - 0.22 0.33 0.44
4 ( gt0.3) - 0.25 0.25 0.50
4 - classes
The overall accuracy with four classes is around
51. If we exclude the values of ATI within the
confidence interval corresponding to the error
measurements of SMC values that can be
misclassified, the accuracy raises to 81.
SMC/ ATI 2 (lt0.055) 3 (0.055-0.085) 4 (gt0.085)
2 (lt0.20) 0.76 0.24 -
3 (0.20-0.30) 0.12 0.60 0.28
4 ( gt0.30) - 0.50 0.50
3 - classes
In this case, the overall accuracy is around 65,
and rises to 88 considering the misclassified
values due to their position very close to the
class boundaries.
12
Analysis of temporal trends France
Comparison of the temporal trend among SMC
(cm3/cm3), ATI originally calculated and ATI
filtered. H stands for High NDVI values (gt 0.4)
and L stand for Low NDVI values (lt 0.4).
No filter Simple filter Filter using AMSRE data
NDVI lt 0.4 0.61 0.68 0.70
NDVI gt 0.4 0.23 0.24 0.43
Temporal comparison between ATI and measured soil
moisture values (SMC) over 1 year period for
France test site. The values represent the
determination coefficients between ATI values and
SMC in the different cases considered.
13
SMC classes from ATI
ATI/SMC 2 (lt0.06) 3 (0.06-0.085) 4 (gt0.085)
2 (lt0.25) 0.78 0.22 -
3 (0.25-0.35) 0.37 0.53 0.10
4 ( gt0.35) 0.17 0.75 0.08
3 - classes
The overall accuracy with three classes is around
58. If we exclude the values of ATI within the
confidence interval corresponding to the error
measurements of SMC values that can be
misclassified, the accuracy raises to 73.
The ranges adopted are slightly different from
the previous test site, because the SMC values
were in general higher while the corresponding
ATI did not change due to the presence of
vegetation detected with high NDVI values. In
fact in the confusion matrix most of the values
of SMC higher than 0.35 cm3/cm3 are in class 3
instead of 4. This happens because all the
highest SMC values were in the period with the
highest values of NDVI.
14
Example of ATI and derived SMC classes
15
Time series on the South Tyrol test site
Matchertal watershed One of the driest valley
in South Tyrol
ATI (K-1)

• The mismatch between ATI values and SMC
  measurements can be due to
• High level of vegetation NDVI is generally
  higher than 0.6
• Difficulty in eliminating completely the cloud
  presence (the completely cloud free images were
  only 20 over 2010 acquisitions)
• Topography effect
• Time of the acquistions

16
Time of acquisitions
17
Spatial analysis

• ATI spatial distribution has been analyzed
  through ANOVA with respect to
• - Land cover (1 urban areas21 agriculture,
  arable land 22 agriculture, vineyards and
  other standing crops
• 31 natural vegetation, trees, forest 32
  natural vegetation, grassland 41 no
  vegetation, rocks 42 no vegetation, debris
• 43 no vegatation, glacier 5 water0 no
  data)
• Elevation classes (gt500m, 500-1000 m, 1000-1500m,
  1500-2000m, gt 2000m)
• NDVI classes (8 classes with border values of
  0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7 and 0.8)

Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects
Dependent Variable145 Dependent Variable145 Dependent Variable145 Dependent Variable145 Dependent Variable145 Dependent Variable145 Dependent Variable145 Dependent Variable145
Source Type III Sum of Squares df Mean Square F Sig. Noncent. Parameter Observed Powerb
Corrected Model .274a 81 .003 12.780 .000 1035.194 1.000
Intercept .509 1 .509 1924.258 .000 1924.258 1.000
LC .023 10 .002 8.819 .000 88.192 1.000
NDVI_145 .014 8 .002 6.651 .000 53.206 1.000
LC NDVI_145 .024 63 .000 1.424 .016 89.743 1.000
Error 1.920 7256 .000
Total 26.702 7338
Corrected Total 2.194 7337
a. R Squared .125 (Adjusted R Squared .115) b. Computed using alpha .05 a. R Squared .125 (Adjusted R Squared .115) b. Computed using alpha .05 a. R Squared .125 (Adjusted R Squared .115) b. Computed using alpha .05 a. R Squared .125 (Adjusted R Squared .115) b. Computed using alpha .05 a. R Squared .125 (Adjusted R Squared .115) b. Computed using alpha .05 a. R Squared .125 (Adjusted R Squared .115) b. Computed using alpha .05 a. R Squared .125 (Adjusted R Squared .115) b. Computed using alpha .05 a. R Squared .125 (Adjusted R Squared .115) b. Computed using alpha .05
Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects Tests of Between-Subjects Effects
Dependent Variable156 Dependent Variable156 Dependent Variable156 Dependent Variable156 Dependent Variable156 Dependent Variable156 Dependent Variable156 Dependent Variable156
Source Type III Sum of Squares df Mean Square F Sig. Noncent. Parameter Observed Powerb
Corrected Model .234a 51 .005 24.856 .000 1267.661 1.000
Intercept 1.284 1 1.284 6955.441 .000 6955.441 1.000
El .011 5 .002 11.751 .000 58.754 1.000
NDVI_156 .043 8 .005 29.079 .000 232.630 1.000
El NDVI_156 .042 38 .001 6.035 .000 229.344 1.000
Error 1.353 7327 .000
Total 20.566 7379
Corrected Total 1.587 7378

NDVI and Land cover classes are significant in
all the analyzed days. This impact (value of
F-statistic) depends on the day, so probably on
NDVI value phenology BUT weather conditions
(cloud cover, fog etc) also might play here
important role!
18
Spatial analysis contd
19
Example of ATI maps
20
Cross-comparison with SEVIRI- MSG data
A cross-comparison with METEOSAT MSG data is
under evaluation. Some MSG images contemporary
to MODIS acquisitions have been processed and
analyzed over the Emilia Romagna test sites. One
of the major problems is the use of a correct
cloud mask in order to not introduce not
corrected ranges of ATI values.
21
Conclusions and future steps

• This work proposes a multi-sensor analysis in
  order to determine soil moisture classes based on
  ATI derived from MODIS images. 3-4 classes of SMC
  were detectable.
• For the approach, main limitation are the cloud
  coverage, the acquisition time of day and night
  images and the presence of vegetation.
• The ATI values have been compared with ground
  measurements and with soil moisture maps
  estimated from SAR images in order to verified
  the spatial pattern consistency. The main
  limitation remains the low repetition time of SAR
  images
• The temporal trends have been filtered by using
  the AMSR-E data in order to reduce the effect due
  to outliers (time of acquisition, clouds, etc..).
  In this case, the main limitation still remains
  the presence of vegetation.
• First comparisons between MODIS and MSG images
  indicate a good correlation between these data.
  For MSG a good cloud cover is highly recommanded.
• Future developments will include
• Further comparison with MSG data, including
  improvement of the cloud mask
• Definition of a calibration procedure between
  MODIS and MSG data
• Integration of MODIS time series with MSG data
• Considering other data such as SMOS, Aquarius

22
Thank you for the attention!Comments/questions?