trishna : une nouvelle mission satellitaire …...jean-pierre lagouarde inra, umr 1391 ispa, 33140...
TRANSCRIPT
Jean-Pierre Lagouarde
INRA, UMR 1391 ISPA, 33140 Villenave d'Ornon, France
TRISHNA : UNE NOUVELLE MISSION SATELLITAIRE À HAUTE RÉSOLUTION SPATIO-TEMPORELLE POUR LE
SUIVI DE L’ENVIRONNEMENT
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
TRISHNA : A NEW HIGH RESOLUTION THERMAL
INFRARED INDO-FRENCH MISSION CONCEPT
FEW Resources.org
General context
Global change (climate, population growth, land use, urbanization, deforestation…) increasing scarcity and deteriorating quality of the water resource monitoring of the water cycle more and more crucial
Thermal infrared → AET, EB (evapotranspiration réelle, bilan d’énergie)
Variability of processes at local scale Need of Spatial systems combining both high spatial resolution and revisit capacities, not existing today
Several projects from the mid 90’s (since IRSUTE, CNES Seguin et al., 1999)…
● 4 LWIR + 3-4 VNIR ● 1 day / 32 km swath● Microbolometers
Lagouarde et al., 2013
…MISTIGRI (CNES)
● HyspIRI (7 LWIR+1MWIR) + 3-4 VNIR● 3 days / Global coverage
● Cooled detector
Crebassol et al., 2014
…THIRSTY (CNES/NASA JPL)
CNES-ISRO projectTRISHNA
Seguin et al. (1999), Rem. Sens. of Environ., 68, 357-369Lagouarde et al. (2013), Int. J. of Remote Sensing, 34 (9-10), 3437-3466Crebassol et al. (2014), IGARSS Symposium, July 13th-18th, Québec City, Canada
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
1) Ecosystem stress and water use design driver 2) Coastal and inland waters design driver
3) Urban4) Solid Earth5) Cryosphere6) Atmosphere
TRISHNA scientific objectives
Thermal infraRed Imaging Satellite for High-resolution Natural resource Assessment
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Part of cultivated area under irrigation
Proportion of total water withdrawal
withdrawn for agriculture
20% arable land irrigated →
45% food production (FAO)
Scientific objectives - Ecosystem stress and water use
70% of world water used for agriculture
Severe water stress in manyparts of the world
Rapid increase of water consumption
Trends in water use by sector (UNEP)(http://www.unep.org/dewa/vitalwater/article43.html)
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
● Agriculture/forestry
- water stress detection / water needs / irrigation optimisation
- water resources management
- growth/ crop production, food security (see 2011 G20)
- impact of agricultural practices on water use, climate change adaptation
- forest fire risks, frost detection…
● Biogeochimical cycles
- carbon cycle ↔ global warming processes
- water quality
- soil pollution
- arctic permafrost
● Hydrology
- link with meteorology (mesoscale circulation)
- water budgets and biogeochimistry at watershed scale
● Ecosystem monitoring, ecology
- microclimates, biodiversity
- natural vegetation droughts
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Scientific objectives - Ecosystem stress and water use
res=6 km
res=0.75 km
Modelled SST fields in the California Upwelling System: zooms of the black box at 6 and 0.75 km resolutions (Capet et al., 2008)
Scientific objectives - Coastal and inland waters
Gulf Stream portion seen by the TIRS sensor, Landsat 8, April 9, 2013
2.5 km diameter eddy adjacent to Catalina Island (CA): airborne SST image (left) within situ temperature transects from two boat tracks (right). (after Holt et al., 2012)
Coastal waters
Submesoscale activity (mixing processes) ↔ ecosystems productivity (phytoplankton)
Gaz fluxes (CO2, CH4) at the air-sea interfaces
Coastal zone monitoring and management: water quality, algaeblooms, halieutic resource, fresh water resurgences, discharges(water, pollutants, thermal plumes…)
10x10km
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Sea ice
Monitoring melting/freezingprocesses (meltponds, leads, polynyas)
Feedbacks climate ↔ melting ice
Thermal ice thickness as derived from AVHRR
surface temperatures (18 Feb 2004, 1200 UTC) in
the eastern Laptev Sea (After Krumpen et al., 2011)
GTN-L lakes (GCOS) [Essential Climate Variable Ts]
Deltas, estuary hydrology, lagunaes
Biological activity and productivity
Warning of water borne diseases (application to human health)
Thermal discharges
…
Inland waters (lakes and rivers)
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Scientific objectives - Coastal and inland waters
● Urban heat island and comfort (welfare of inhabitants, health…)
● Urban vegetation
● Urban and peri-urban hydrology (run-off, urban planning)
● Urban meteorology and atmospheric flow (dispersion of pollutants…)
● Anthropogenic fluxes and energy consumptions (heating, air
conditionning …) AVHRR Aug. 9, 2003 UHI
over Paris (Dousset, 2007)
2008 2050
3.3 5
Urban population (109 inhabitants)
URBAN
● Monitoring volcanic activity(prediction of eruptions, mitigation of risks…)
● Detection of peat and coal fires(pollution,CO2…)
● Detection of thermal anomalies
(geothermal exploration, earthquakes precursors...) Thermal precursors to lava flow at Kliuchevskoi:
anomalies in the crater, ASTER data (Murphy et al., 2013)
SOLID EARTH
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Scientific objectives - Secondary objectives
CRYOSPHERE
ATMOSPHERE
(from NASA Earth Observatory)
Himalaya-Karakoram-Hindukush (HKH), virtually a ‘third pole’
● snow and glacier melt runoff
● debris detection in glaciers
● dynamics of glacial lakes
● surface radiation budget
● clouds (high clouds, cirrus…)
● atmospheric water vapor
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Scientific objectives - Secondary objectives
0verview on recent research for defining mission specifications
• Impact of atmospheric turbulence on LST measurementsaccuracy
• Analysis of TIR directionnal anisotropy• Development of models and methodologies for estimating
surface fluxes
Recent attention paid on particular topics :
… reveals crucial for determining the mission specifications
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Impact of atmospheric turbulence on LST measurements
Experimental measurements of LST temporal fluctuations over a dry maize field
0 200 400 600 800 1000 120033
34
35
36
37
38
39
time (sec)LS
T (
°C)
0 200 400 600 800 1000 120033
34
35
36
37
38
39
time (sec)
LS
T (
°C)
0 200 400 600 800 1000 120033
34
35
36
37
38
39
time (sec)
LS
T (
°C)
0 200 400 600 800 1000 120033
34
35
36
37
38
39
time (sec)
LS
T (
°C)
95 m50 m
20 m5 m
6°C
6°C
Similar fluctuations are observed over a bare dry smooth soil
Low frequency ↔ Planetary BL (PBL), high frequency ↔ Surface BL (SBL) 9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Statistics of the satellite LST departure from its mean value
0
10
20
30
40
50
60
70
80
90
100
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2
Fre
qu
en
cy (
%)
Ts uncertainty (°C)
Seq 1
Seq 2
maize
bare soil
Bray 1995
Experimental characterization of the uncertatinty on Ts
0
10
20
30
40
50
60
70
80
90
100
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6
Fre
qu
en
cy (
%)
Ts uncertainty (°C)
7m
21m
35m
56m
105m
203m
LES simulation, Pine forest
0.6 °C for ~60% measurements0.8 °C for ~80% measurements
at 50m resolution
Lagouarde et al., RSE 2013Lagouarde et al., RSE 2015
Flexibility on NeDTspecification?
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Impact of atmospheric turbulence on LST measurements
UrbanImpact of rows
(vineyard)
Lagouarde et al., RSE 2004, 2010 Lagouarde et al., IEEE, 2013
Lagouarde et al., RSE 2000
Impact of pine stand structure
20°
30°
40°
50°
60°
30°
60°
90°
120°
150°
180°210°
240°
270°
300°
330°
-8
-6
-4
-2
0
2
4
20°
30°
40°
50°
60°
30°
60°
90°
120°
150°
180°210°
240°
270°
300°
330°
-6
-4
-2
0
2
4
6
~ 11:30 UT
~ 14:00 UT
-12
-10
-8
-6
-4
-2
0
2
4
6
-12
-12
-12
-12
-12
-12
-12
-12-10
-10
-10
-10
-10
-10
-10
-10
-10-8
-8-8
-8 -8
-8
-8
-8
-8
-6
-6
-6
-6
-6-6
-6
-6
-4
-4
-4
-4
-4
-4
-4
-2
-2
-2
-2
-2
-2
0
0
0
0
0
0
22
2
2
2
4
4
6
10° 20° 30° 40° 50°
30°
60°
90°
120°
150°
180°
210°
240°
270°
300°
330°
360°
-14
-12
-10
-8
-6
-4
-2
0
2
-14
-14
-14
-14 -14
-14
-14
-14
-12
-12
-12
-12
-12
-12
-12
-12-10
-10
-10
-10
-10
-10
-10
-10-8
-8
-8
-8
-8
-8
-8
-6
-6
-6
-6
-6
-6
-6
-4
-4
-4
-4
-4
-4
-2
-2
-2
-2
-2
0
0 0
2
10° 20° 30° 40° 50°
30°
60°
90°
120°
150°
180°
210°
240°
270°
300°
330°
360°
E - W
N - S
48 ans
15 ans
6 ans
Directional anisotropy
LST prone to important directional anisotropy effects and hot spot
-10
8
6
0
2
Dir
ecti
on
alan
iso
tro
py
(°C
)
-2
4
-8
-6
-4
VZA
Impact of structure(SCOPE simulated)
Comparison of HS inprincipal solar planefor 3 crop structures(spherical, planophile,erectophile)
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Directional anisotropy (modelling)
Measured ModeledU
rban
Mar
itim
e P
ine
stan
dModelling directionalanisotropy:
3D models (urban daytimeand nighttime[Lagouarde et al., RSE 2010 and 2012]
SCOPE multilayer model [Duffour et al., 2015 and 2016a]
Parametric approach[Duffour et al., 2016b]
Duffour et al. / RSE, 158 (2015) 362–375Duffour et al. / RSE, 177 (2016a) 248–264Duffour et al. / RSE, 186 (2016b) 250–261
Research currently conducted toimprove parametric approaches bycombining TIR with VNIR anisotropy→ See Zunjian Bian’s Poster
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Modelling surface fluxes
Different approaches are followed within the TRISHNA SWG group :
Modèle mono-source Modèle double-source
Ts Tsol
rara
Ta Ta
T0 T0
re
rac
ras
1 source- and 2 source-models based on resistive schemes (Kustas et al.)
SEBAL (Bastiansen, 1998)
B
A
Hmax
LEmax
Ts
Tem
pe
ratu
re d
e s
urf
ace
Albédo
S-SEBI (Roerink et al., 2000)
EF = CB / AB
LE = EF*AE
LE : flux de chaleur
latente
EF : fraction
évaporative
AE: énergie
disponible
WDI (Water Deficit Index) (Moran et al 1994)
METRIC (Allen et al., 2007)
Ta
ux
de
co
uv
ert
ure
vé
gé
tale
=f(
ND
VI)
WDI = AC / AB
WDI = 1 – LE / LEp
LE : flux de chaleur
latente
WDI: indice de stress
LEp : LE en
conditions
potentielles
C
Ts - Ta
BA
C
Contextual approaches (S-SEBI, WDI,…)
Lagouarde and Boulet, 2016. Land Surface Remote Sensing in Continental Hydrology (chapter 10)
Multi-layers TSVA models (MuSICA, SCOPE…) Ogée et al., 2003; Van der Tol et al., 2009; Duffouret al, 2015
soil
canopy
atmosphere
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Important methodological know-how → Efforts to develop practical tools for ET mapping
s
From Richard Allen (METRIC/SEBAL workshop Fort Collins, Colorado, February 7, 2005)
Operational applications to water management developed in USA
Development of EVASPA : a tool for
mapping actual evapotranspiration
Gallego Elvira et al., DOI: 10.1016/j.proenv.2013.06.035
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Modelling surface fluxes
Regional-scale ET and agricultural consumptive water use from Indian Geostationary Satellites
In : Bhattacharya et al, 2013; ISRO-GBP Scientific report
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Main mission specifications
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
MISTIGRI
HyspIRI
TIRS Landsat 8 TM LandsatLandsat ETM+
m20 30 50 7040 8060 90 100 110 120
ASTER
Mission specifications - Spatial resolution
Acces to field scale require <100m
(Agam 2007, Garrigues 2006, Kustas2004….)
SBL Turbulence limitsaccuracy
0 200 400 600 800 1000 120041
42
43
44
45
46
47Ts(t) pour la résolution de 8 m
Temps (s)
Tem
péra
ture
(°C
)
Puducherry region
Optimum range of resolution at nadir
Degradation of end of swath pixels
Nadir resolution :
TRISHNA
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Revisit imposed by 4 combined constraints:
0 50 100 150 200 250 300 350 4000
0.5
1
1.5
2
2.5
3
3.5
4
4.5
DoY
Num
be
r clo
udfr
ee
da
ys
/ 5
da
y p
eri
od
AVIGNON
1 day
2 days
5 days Statistical analysis of ahourly solar radiationdataset (1993-2009, INRAAGROCLIM) over France
Cloud frequency analysis withcriteria: 1 cloudfree day / 5 daysrequired], RAQRS 2010
Analysis of 2000-2009 MODIS cloud masks
1 day 2 days
● minimum availability of cloud free data1
● providing sufficient accuracy on derived products (meteo, irrigation…)
Analysis of ISBA-AGsETR (IPCC scenario A1B,period 1950 - 2100 )
Delogu, E. et al. (2012). HESS, 9, 1699-1740
2
DoY
Flux, LST
0 2 4 6 8 10 12 14 16 18 2034
34.5
35
35.5
36
36.5
37
37.5
38maïs sec
time (minutes)
Ts (°
C)
4°C
20 mn
Dry maize
Assimilation and/or interpolations more robust with higherrevisit
● Necessity of reducing the impact of atmospheric turbulence uncertainty
3
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mission specifications - Revisit
● global coverage requirement orbit/revisit/swath/scan angle4
THIRSTY666 km orbit
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mission specifications - Revisit
Also refer to Anderson et al., 2012
uncertainty SBL (HF)turbulence
Revisit1 2 3 54
MISTIGRI HyspIRI
days
THIRSTY
Temporal variability of meteoforcingwater status (rainfall, irrig.)
Cloud frequencylimits the number of observations
Global coverageToo large scan angles
TRISHNA
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mission specifications - Revisit
1.0
0.9
0.8
0.7
0.5
0.4
0.6
1413121110
Revisit time (LST)
Nas
h e
ffic
ien
cy
1 Overpass time must cope with maximum sensitivity for flux estimation
→ ~12:00 - 13:00 UTC overpass
time providing maximum accuracy
on fluxes retrieval
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mission specifications - Overpass time
Minimizing the sensitivity of LST measurement to overpass time
-5
0
5
10
15
20
25
30
35
40
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20Su
rfac
e t
em
pe
ratu
re (
°C)
Time(UT)
223
252
257
268
272
273
274
275
276
277
286
289
294
296
313
LST at different dates along the year (pine stand)
→ 13 LST minimises possible errors due to variation in acquisition time
→ 13 LST more suited to models having30mn-1hour time steps
2
~
4°C/h
1.0
0.9
0.8
0.7
0.5
0.4
0.6
1413121110
Revisit time (LST)
Nas
h e
ffic
ien
cy
1 Overpass time must cope with maximum sensitivity for flux estimation
→ ~12:00 - 13:00 UTC overpass
time providing maximum accuracy
on fluxes retrieval
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mission specifications - Overpass time
13:00 LST daytime 01:00 nightime overpass
1:00 AM preferred for ST measurements over water surfaces (by avoiding remainingskin thermal effects)
3 SST measurements
simulation 13:00 TU 15/07/2004 Toulouse
20°
30°
40°
50°
60°
30°
60°
90°
120°
150°
180°210°
240°
270°
300°
330°
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
For mid-latitudes, scan perpendicular to principal plane →limitations of hot spot effects
4 Hot spot impact
!Not true for inter-tropicalzone!
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mission specifications - Overpass time
0
10
20
30
40
50
60
70
80
90
100
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 2
Fre
qu
en
cy (
%)
Ts uncertainty (°C)
Seq 1
Seq 2
maize
bare soil
Atmospheric turbulence- induceduncertainty on LST
NeDT specification betterthan 0.3 K not necessary forcontinental surfaces at 50 mresolution
TBC for water surfaces
0 200 400 600 800 1000 120041
42
43
44
45
46
47Ts(t) pour la résolution de 8 m
Temps (s)
Tem
péra
ture
(°C
)
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mission specifications - Instrument sensitivity
The TIR Hot Spot within the inter-tropical area
No robust model of HotSpot available today large uncertainty of LSTclose to HS
-10
8
6
0
2
Dir
ect
ion
alan
iso
tro
py
(°C
)
-2
4
-8
-6
-4
VZA
HS from SCOPE simulations(Duffour et al., RSE 177, 248-264, 2016)
Duffour et al. / RSE 186 (2016) 250–261Duffour et al. / RSE 177 (2016) 248–264
Debate between (i) a 3 day (666 km) and (ii) a 8 day(679 km) orbit providing 3/2/3 day sub-cycles withdifferent viewing configurations
Scan = HS
January February March
April JuneMay
July August September
October DecemberNovember
Bangalore : position of the TIR hot spot throughout the yearRL model simulations (qualitative)
In inter-tropical (IT)area, HS very oftenclose to the scanline
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mission specifications - Orbit definition
Simulations being performed with a end-to-end simulator to consolidateTIR bands (jointly at CNES, ONERA, IRD and Univ. Valencia)
TOA Luminance at satellite level
Instrument filters and noise
Luminance in satellite bands
T, separation
PrescribedLST and () Estimated and LST
?Comparison/evaluation
Atmospheremodel
Atmospherecorrection
(SW)
4 bands : 8.6 µm = 0.35 µm9.1 µm = 0.35 µm
10.3 µm = 1 µm11.5 µm = 1 µm
0.0
0.2
0.4
0.6
0.8
1.0
7 8 9 10 11 12 13 14 15
Wavelength (m)
Atm
osp
heri
c T
ran
sm
issiv
ity
spectral
smoothed
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
7 8 9 10 11 12 13 14
Wavelength (m)
Em
iss
ivit
y
MEAN
MAX
MIN
MEAN-STD
MEAN+STD
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mission specifications - TIR spectral bands specification
→ Registrating TIR imagery→ Cloud detection and filtering→ Disaggregation of TIR→ Contextual approaches→ Surface parameters estimation→ …
VNIR bands together with TIR for :
Simultaneity between TIR and VNIR measurements needed both instruments on the same platform
Band Central wavelength (µm)
Need Status
Blue 0.485 Cloudsdiscrimination
Mandatory(degraded resol)
Green 0.555 Coastal, sediments,snow
Mandatory
Red 0.650 Vegetation Mandatory
Near Infrared 0.860 Vegetation Mandatory
SWIR 2 2.13 Atmosphere ISRO requirement
SWIR 1 * 1.38 Atmosphere, cirrus detection
Optional, TBD(degraded resol)
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mission specifications - VNIR spectral bands specification
Current studies - Concept & technologies trade-offs
Concept
Push-broom concept killed off
One study on a scanner
One study with step and stare
Technologies
µbolometer too “slow” for such required performances
MCT cooled detector as
baseline
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU
Mislestones
• November 2017 : French-Indian Mission Design Review, to verify :• the scientific needs and how they are interpreted in a mission specification
• the data distribution policy
• The work share between France and India
• 2018-2019 : phase A, including an industrial instrumental phase A study
• Launch 2024-2025 ?
9ème Journée9ème Journée Thématique de l ’OASU, Bordeaux, 12 octobre 2017SU