DROUGHE ESTIMATION AND MAPPING - CESBIO Mehrez ZRIBI, Michel le page - theia-land.fr
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
DROUGHE ESTIMATION AND
MAPPING
Mehrez ZRIBI, Michel le page
CESBIO
CESBIO/TOULOUSE
• Toulouse: second university city in France
• The world's largest higher education
institution in aerospace engineering
• Strong presence of the aeronautics, space
and high technology industry
• Toulouse III is 6th in the world in the
Shanghai ranking in remote sensing
• The climate and weather are super pleasant
2
Land surface Surface-climat Remote sensing
modelling
Water Agro-
ecosystems
In situ
observations
3
CESBIO and partners observatories
South West: OSR
India: IRP CEFIRSE
Lebanon: GDRI OLIFE
Morocco: LMI TREMA
Tunisia: LMI NAILA
4
ICOS, JECAM, GEOGLAM, FLUXNET, ZA PYGAR, IR OZCAR
Satellite missions in CESBIO
2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032
SMOS (ESA CNES)
VENµS (CNES, IAI)
Biomass (ESA), P-Band
Color codes
Altimeters TRISHNA (CNES/ISRO)
L-Band Passive
Optical
Thermal
Radar
Precipitation SMOS-HR
5
Satellite products
6
Spatialized plateforms
29 Juin 2006 Model de transfert radiatif Cartes de Green Area Index
(e.g. BV-net de l'INRA Avignon) m2 feuille/m2 sol
Spot, Landsat,
Sentinel 2 29/06/2006
Initialisation, NDVI 26/07/2006
Forçage, 09/09/2006
assimilation
calibration
Modèle
Sentinel 1, Terrasar X spatialisé
Cartes dynamiques
des OSCartes
et des
occupation
pratiques SAFYE/ SAFY
du sol FAO56 Rendement indice foliaire
culturales
Eau
8 km CO2
Ré-analyses
Météo SAFYE-CO2
Exemple de sortie
Validation par stat. régionales
Cartes des (Rendement / Irrigation)
sols
Validation par stations de
mesure des flux (ET, Humidité Bilans C des parcelles de blé en
2011 (gC.m-2)
sol…)
7
Impacts of drought
• Economic
• Social
• environmental
• Impacts increase in response to increasing vulnerability
resulting from increased pressure on limited water
resources, population growth and many other factors.
• Post-crash response increases vulnerability.
• The impacts differ from one country to another.
Drought Characteristics
• Normal component of climate variability
• No universal definition
• Complex
• Interdisciplinary
• Impacts can be economic, social, environmental
• Impacts can persist for years
❖ The beginning and end of the drought are difficult to
determine
❖ No precise and universally accepted definition of drought
❖ Non-structural impacts and spread over a large
geographical area
Agricultural Drought
→ Meteorological drought affecting agriculture
→ Usually, the first economic sector to be affected
→ Shortage of precipitation, ET, soil moisture, etc.
→ Demand for water from the plant in relation to
available soil moistureMonitoring of Drought • Importance of drought indices → Simplify complex inter-relationships and provide a good communication tool for many audiences → Quantitative assessment of climatic conditions → Provide a historical perspective that can be used in planning and design applications
Data and Methods
Long multi-spectral and multi-
resolution time series
SpotVeg, Proba-v – Sentinel-3
MODIS - VIIRS
AVHRR
ERS
ASCAT (Metop A-C)
SMOS, SMAP
1980 1990 2000 2010 2020Long multi-spectral and multi-
resolution time series
SpotVeg, Proba-v – Sentinel-3
MODIS - VIIRS
AVHRR
ERS
ASCAT (Metop A-C)
SMOS, SMAP
1980 1990 2000 2010 2020Selected Indices
• MODIS NDVI : Vegetation vigor
• MODIS LST : Land Surface Temperature
• ASCAT SWI : Soil Moisture
m (
s i
t )e − (t −ti ) T
SWI (t ) = i
for ti t
e
i
− (t −ti ) TNormalization of Remote Sensing Obs. VAI: Vegetation Anomaly Index NDVI: NDVI at one date NDVImean: mean of NDVI for a selected period s: standard deviation of NDVI → The same normalization is applied to Soil Moisture (MAI) and Temperature (TAI)
Mixed indices
Indi = αi VAIi + βi MAIi
Indi − Indi mean
GDIi =
σInd,i
Zribi et al., RS, 2021Classify drought on two dimensions
VAI Ambiguous: Healthy
vegetation but Dry Soil
DRY
-Inf -2S -1S 0 1S 2S Inf
-Inf
-2S
8 9
-1S 4 5
MAI
0 1
Ambiguous: Low 1S 7 6
vegetation but Wet
Soil 11 10 WET
2S
Inf
Le Page, Zribi, Scientific Reports, 2019Search for similar years using a drought vector
Study Area: North-West Africa ESA CCI-LC V2.0.7 GLCNMO V3.1
Caracterization of agricultural calendars
Minimum of NDVI Maximum of NDVI
Average Date of maxNDVI Stdev of Date of maxNDVIOccurrences of Situations
VAI
-Inf -2S -1S 0 1S 2S Inf
-Inf
-2S
8 9
-1S 4 5
MAI
0 1
1S 7 6
11 10
2S
Inf
2007 to 2018, November to AprilOccurrences of Situations: Normal and moderate 2007 to 2018, November to April
Occurrences of Situations: Extremes 2007 to 2018, November to April
Occurrences of Situations: Ambiguous 2007 to 2018, November to April
Drought qualification for the study area
50
VAI
45
% Moderately Dry
-Inf
-Inf -2S -1S 0 1S 2S Inf 40
35 %Very Dry
-2S
8 9
-1S 4 5 30
25
%
MAI
0 1
1S
20
7 6
11 10 15
2S
Inf
10
5
0
Avg.
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
25
Good Vgt but Dry soil Good Vgt but Dry soil 2
VAI
-Inf -2S -1S 0 1S 2S Inf
Low Vegt but Wet soil 2 Low Vegt but Wet soil
-Inf 20
-2S
8 9
-1S 4 5 15
MAI
%
0 1
1S 7 6 10
11 10
2S
Inf 5
0
Avg.
2008
2009
2010
2012
2013
2014
2017
2018
2011
2015
2016
2007 to 2018, 90 administrative areas. November to April for each agricultural yearDrought qualification for the study area
50
45
% Moderately Dry
40
35 %Very Dry
30
25
% 20
15
10
5
0
Avg.
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
25
Good Vgt but Dry soil Good Vgt but Dry soil 2
Low Vegt but Wet soil 2 Low Vegt but Wet soil
20
15
%
10
5
0
Avg.
2008
2009
2010
2012
2013
2014
2017
2018
2011
2015
2016
2007 to 2017, November to April for each agricultural yearDrought qualification for the study area
50
45
% Moderately Dry
40
35 %Very Dry
30
25
% 20
15
10
5
0
Avg.
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
25
Good Vgt but Dry soil Good Vgt but Dry soil 2
Low Vegt but Wet soil 2 Low Vegt but Wet soil
20
15
%
10
5
0
Avg.
2008
2009
2010
2012
2013
2014
2017
2018
2011
2015
2016
2007 to 2017, November to April for each agricultural year50
45
% Moderately Dry
40
35 %Very Dry
Use cases for the peculiar years
30
25
%
20
15
10
5
0
Avg.
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
25
Good Vgt but Dry soil Good Vgt but Dry soil 2
Low Vegt but Wet soil 2 Low Vegt but Wet soil
20
15
%
10
5
0
Settat, Morocco Aïn Defla, Algeria Kairouan, Tunisia
Avg.
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2007-08
2015-16
2017-18
NDVI SWI Cumulative RainfallPerformance of the forecast
Percentage of Percentage of Average Rank
First Analog First or Second
Analog
November 21.2 38.4 4.09
December 32.6 53.0 3.12
January 47.5 72.4 2.46
February 57.3 80.0 2.03
March 71.4 91.7 1.51
April 100 100 1Use Case: Settat, Morocco
Use Case: Settat, Morocco
1 1 -
3 1 1
3 1 1
2 1 3
1 1 1A Webapp with monthly updates
Thank you for your attention
You can also read
