flood and drought monitoring and prediction by satellite ... · flood and drought monitoring and...
TRANSCRIPT
Flood and Drought Monitoring and Prediction by Satellite-based Microwave Remote Sensing
Regional Training Workshop "Advances in Remote Sensing Application in Water Resources Management"
Toshio KoikeDirector, International Centre for Water Hazard and Risk Management (ICHARM)
Council Member, Science Council of Japan (SCJ), Cabinet Office of Japan Professor Emeritus, the University of Tokyo
Chair, River Council of Japan
Definition of Remote Sensing
Remote Sensing is a technology for identifying a target and estimating its physical, chemical and biological conditions without touching by using its inherent characteristics of emission, reflection, absorption and transmission of electromagnetic wave and its radiation transfer.
MWIRVISUV frequency(Hz)
0.7um-1mm0.4-0.7um 1mm-1m
WATER
StrongDipoleMolecule
Unique Roles in the Earth Environment• Large Specific Heat of Liquid Water: Ocean as a heat transporter
• Large Heat Exchange through Liquid – Gas Phase Transition Water vapor as a heat transporter
• Solid ICE Crystal Lattice: Ice floats in water & bine drives
the great ocean conveyor belt.
OHH
HH
OO
HH- - -
+++
+++
- - -Strong Dielectric Material
ALOS-2 GPM-Core
+++
- - -
GCOM-W1
Active Passive
Microwave Remote Sensing
5
Soil Moisture Vegetation/Snow
Active & PassiveMicrowave Sensors
Precipitation
Surface Emissivity & Temp.
Microwave Remote Sensing of Land Hydrology
ALOS-2
GPM-Core
GCOM-W1
6
Satellite Precipitation Sensors
Land Surface Monitoring by SARSAR measures a intensity of microwaves reflected from the earth’s surface
The intensity is called “backscattering coefficient σ0 = f( Mv, Sd, Cl, Sv, D )
Soil moisture:Mv : Volume fraction of soil moisture (vol.)
Surface roughness:Sd : Standard deviation of surface height (cm)Cl : Surface correlation length (cm)
Soil parameter:Sv : Volume fraction of soil grains (vol.)D : mean diameter of soil grains (cm)
Surface scattering
volume scattering
Surface scattering
volume scattering
Scattering image
Nigeria Flood in September 2018
https://edition.cnn.com/2018/09/18/africa/nigeria‐flood‐national‐disaster/index.html
https://www.aljazeera.com/news/2018/09/nigeria‐floods‐kill‐100‐people‐10‐states‐180917193612830.html
Identify inundation area on Sep‐22, 2018MODIS (optical)MODIS (optical) Sentinel1 (SAR)Sentinel1 (SAR)
Flood Area mapFlood Area mapFlood Area by SARFlood Area by SAR
Source of back image: "NASA Worldview"
Source:Copernicus Sentinel Data
Target area
Easily identification & high frequency but covered with cloud
All weather & high spatial resolution but low frequency
Niger River Flood Area Map
1) Sep‐04, 20181) Sep‐04, 2018 2) Sep‐16, 20182) Sep‐16, 2018 3) Sep‐22, 20183) Sep‐22, 2018 4) Sep‐28, 20184) Sep‐28, 2018
Source of back image: "NASA Worldview"
5) Oct‐10, 20185) Oct‐10, 2018 6) Oct‐22, 20186) Oct‐22, 2018 7) Nov‐03, 20187) Nov‐03, 2018 8) Nov‐15, 20188) Nov‐15, 2018
Physical Measurement Approach
O UB
O MG
O CYMONGOLIA SA
CHINA
RUSSIA
0
1
2
34
5
6
7
8
A B C D E F G H I J
AWS
ASSH
Location of AWS and ASSH in AMSR experimental fields of the study area (UB:Ulaanbaatar, CY:Choir, MG:Mandalgobi, SA:Study area)
Temporal Variation of Spatially Averaged Validation
Seasonal Variation of the Soil Moisture in the Tibetan Plateu
17
Soil Moisture Vegetation/Snow
Precipitation
Surface Emissivity & Temp.
Data Assimilation Coupled with Microwave Remote Sensing
Land Surface Scheme
Vegetation/Snow Model
CloudPhysicsModel
Microwave Remote Sensing of Land Hydrology
ALOS-2
GPM-Core
GCOM-W1
Active & PassiveMicrowave Sensors
observationRadar-AMeDAS
predictionJMA-MSM
uncertainty inthe location of
rainfall systems
Coupled Atmosphere-LandData Assimilation System (CALDAS)
Low freq.LDAS
↓
Soil Moist.Suf. Temp.
↓
High freq.ADAS
Seto, Koike, et al.JGR(2013)
The grant which financed this Pilot for Agriculture Drought Monitoring and Prediction in Brazil was received under the Japan-Bank Program for Main-streaming DRM in Developing Countries which is financed by the Government of Japan
Background:ScientificContribution
+
Dynamic Vegetation Model
AMSR-E AMSR2
Yang, Koike, et al. JMSJ (2007)
Electronic-MagneticWave
Sawada & Koike, JGR (2014)
Data AssimilationCoupled
Natural climate variability
Precipitation deficiency, high temperature etc…
Soil water deficiency
Plant water stress, reduced biomass and yield
Reduced stream flow, inflow to reservoirs, Groundwater deficiency
Economic, Social, and Environmental impacts
Met
eoro
logi
cal
Ecol
ogic
al &
Ag
ricul
tura
lH
ydro
logi
cal
Relationship between ecological and hydrological processes is important for analyzing drought process.
Water-Energy BudgetHydrological Model
(WEB-DHM)
River DischargeSoil MoistureGround WaterDam Storage
Crop Production
Real Time Data Management System
DIAS Archives
NASA GLDAS Optimized LDAS ParametersAMSR2MODIS
GFDLAPCC
Hydromet Data Satellite Seasonal Forecast Global CLVDAS Outputs
Optimized RTM Parameters
NCDC Global Met
JMA Reanalysis
LDAS Reanalysis Statistics
Hydrometeorology‐Agriculture Droughts Prediction System
Satellite-based Land Data Assimilation
(CLVDAS)
Drought analysisWheat production
LAI anomaly from CLVDAS
2007 Morocco Drought
Morocco Algeria Tunisia
from Sawada & Ikoma
Prediction
Water-Energy BudgetHydrological Model
(WEB-DHM)
River DischargeSoil MoistureGround WaterDam Storage
Crop Production
Satellite-based Land Data Assimilation
(CLVDAS)
Real Time Data Management System
DIAS Archives
NASA GLDAS Optimized LDAS ParametersAMSR2MODIS
GFDLAPCC
Hydromet Data Satellite Seasonal Forecast Global CLVDAS Outputs
Ensemble Rainfall Prediction
Ensemble Drought Prediction
Optimized RTM Parameters
NCDC Global Met
JMA Reanalysis
LDAS Reanalysis Statistics
River DischargeSoil Moisture, Ground Water
Dam StorageCrop Production
Rainfall Pattern 1Rainfall Pattern 2
Rainfall Pattern n
Prediction Accuracy EvaluationBias Correction
Weighting
Hydrometeorology‐Agriculture Droughts Prediction System
Agricultural DroughtMonitoring-Prediction
North AfricaAqua AMSR-E
The grant which financed this Pilot for Agriculture Drought Monitoring and Prediction in Brazil was received under the Japan-Bank Program for Main-streaming DRM in Developing Countries which is financed by the Government of Japan
Marengo et al., 2017
Rainfall Anomaly
Number of DryDays
Vegetation WaterSupply Index(VWSI) Anomaly
2012 2013 2014 2015 2016
Background: Long-term Serious Droughts
The grant which financed this Pilot for Agriculture Drought Monitoring and Prediction in Brazil was received under the Japan-Bank Program for Main-streaming DRM in Developing Countries which is financed by the Government of Japan
It is virtually certain that drought will become more severe.
0
200
400
600
800
1000
rainfall mm/yea
r
Annual Average Rainfall
1986‐20002051‐2065
0
200
400
600
800
1000
rainfall mm/yea
r
Annual Average Rainfall1986‐20002051‐2065
AIN BEYA OUEDKALAAT ESSENAM
Mejerda River
29
31
m
00.40.81.21.62
0
2
4
6 Bir Lakhdar Souk Sebt 2
cm
Groundwater Level Variations (satellite-observed)
Groundwater Level Variations (ground-observed)
-10-505
10
2002 2003 2004 2005 2006 2007 2008
Future GW Variations
‐1.3mm/year
GW
sto
rage
ano
mal
y (c
m)
18/20G
W s
tora
ge a
nom
aly
(cm
)GW (TWS – SM)GW (LDAS-WEBDHM)Inter‐comparison
33
Soil Moisture Vegetation/Snow
Precipitation
Surface Emissivity & Temp.
Data Assimilation Coupled with Microwave Remote Sensing
Land Surface Scheme
Vegetation/Snow Model
CloudPhysicsModel
Microwave Remote Sensing of Land Hydrology
ALOS-2
GPM-Core
GCOM-W1
Active & PassiveMicrowave Sensors
Preliminary outputsGautemala
Lake Atescata
LAI output from CLVDAS
2003 2004 2005 2006
2007 2008 2009 2010
2013 2014 2015 2016
2017 2018November 1
Gautemala spatial average
FAOSTAT
FAOSTAT
CLVDAS LAI and major products
3.3. Application: Horn of Africa drought
[FAO, 2011]
[Anderson et al., 2012]
We cannot have the access to many ground observations to develop the drought prediction system.
Leaf Area Index timeseriesC
LVD
AS(R
eal P
redi
cion
)Blue: PredictionGreen: Horn of Africa drought (reanalysis)
[Sawada and Koike, JGR-A, submitted]
Predictions: starting from 1 Oct 2010
CLV
DAS
(Rea
l Pre
dici
on) Predictions: starting from 1 Jan 2011
CLV
DAS
(Rea
l Pre
dici
on) Predictions: starting from 1 Mar 2011
The grant which financed this Pilot for Agriculture Drought Monitoring and Prediction in Brazil was received under the Japan-Bank Program for Main-streaming DRM in Developing Countries which is financed by the Government of Japan
Bias Corrected Rainfall & Hydro-met Forcing
Previous Year Hydro-met Forcing with
Modified Rainfall by Seasonal Prediction
Dec.1
Dec.15
Jan.1
Jan.15
Feb.1
Feb.15
Mar.1
Mar.15
Dec.Dec.1-31
Dec.1-31Jan.-Mar.
Jan. Feb. Mar.
Jan. Feb. Mar. Apr.
Feb. Mar. Apr. M
Nov. Dec. Jan. Feb.
Jan.1-31
Jan.1-31Feb.-Apr.
Feb.1-31
Feb.1-31Mar.-May
Monitoring Prediction
The grant which financed this Pilot for Agriculture Drought Monitoring and Prediction in Brazil was received under the Japan-Bank Program for Main-streaming DRM in Developing Countries which is financed by the Government of Japan
LAI deviation
Crop yield under irrigation condition
LAI deviation
Crop yield under rain-fed condition
DIASdrought system
DIASdrought system
Precipitation
Irrigation water volume and its supply timing
Irrigated area
Irrigated area, water volume and its supply timing: If we obtain the irrigated area, water volume and its supply timing, we can calculate LAI and crop yield under the irrigation condition by inputting precipitation and irrigation water into the DIAS drought system and can compare it with the crop yield under rain-fed condition. In consequence, we can evaluate the effectiveness of irrigation.
Effectiveness of irrigation
Input
Input
Output
Output
Application to estimate required irrigation water volume without the irrigated water volume and its supply timing
Precipitation + Irrigation water
Estimation of the required irrigation water volume: We assumed the averaged crop yield in the period between 2003 and 2017 as the target crop yield of each farmer. The required irrigation water volume was estimated by comparing the target crop yield and the relationship between assumed irrigation water and estimated crop yield. In consequence, our simulations clearly show consistent decreases in required irrigation water volume as the drought condition improved gradually in 2015 and 2016 after a severe drought in 2014.
0.100.150.200.250.300.350.40
0 1 2 3 4 5 6 7 8
P cas
hew
[t/ha
]
Irrigation [Ir, mm/day]
2014
0.100.150.200.250.300.350.40
0 1 2 3 4 5 6 7 8Irrigation [Ir, mm/day]
2015
0.100.150.200.250.300.350.40
0 1 2 3 4 5 6 7 8Irrigation [Ir, mm/day]
2016
1.94 mm/day 1.45 mm/day 0.16 mm/day
0.20
0.25
0.30
0.35
0.40
0 1 2 3 4 5 6 7 8
P bea
ns[t/
ha]
Irrigation [Ir, mm/day]
2014
0.20
0.25
0.30
0.35
0.40
0 1 2 3 4 5 6 7 8Irrigation [Ir, mm/day]
2015
0.20
0.25
0.30
0.35
0.40
0 1 2 3 4 5 6 7 8Irrigation [Ir, mm/day]
2016
1.86 mm/day 1.38 mm/day 0.06 mm/day
56789
1011
0 1 2 3 4 5 6 7 8
P cas
sava
[t/ha
]
Irrigation [Ir, mm/day]
2014
56789
101112
0 1 2 3 4 5 6 7 8Irrigation [Ir, mm/day]
2015
56789
1011
0 1 2 3 4 5 6 7 8Irrigation [Ir, mm/day]
2016
2.70 mm/day 2.10 mm/day1.07 mm/day
0.20
0.40
0.60
0.80
1.00
1.20
0 1 2 3 4 5 6 7 8
P cor
n[t/
ha]
Irrigation [Ir, mm/day]
2014
0.20
0.40
0.60
0.80
1.00
1.20
0 1 2 3 4 5 6 7 8Irrigation [Ir, mm/day]
2015
0.20
0.40
0.60
0.80
1.00
1.20
0 1 2 3 4 5 6 7 8Irrigation [Ir, mm/day]
2016
2.02 mm/day 1.52 mm/day 0.25 mm/day
Cashew nuts
Fejonbeans
Cassava
Corn