Impact of atmospheric turbulence on the accuracy of LST measurements

HAL Id: hal-02795610

https://hal.inrae.fr/hal-02795610

Submitted on 5 Jun 2020

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Impact of atmospheric turbulence on the accuracy of LST measurements

Jean-Pierre Lagouarde, Mark Irvine, Sylvain Dupont

To cite this version:

Jean-Pierre Lagouarde, Mark Irvine, Sylvain Dupont. Impact of atmospheric turbulence on the accu- racy of LST measurements. 3. GlobTemperature User Consultation Meeting, European Space Agency (ESA). INT., Jun 2015, Reading, United Kingdom. n.p. �hal-02795610�

INRA

UMR 1391 ISPA

F-33140 Villenave d'Ornon

lagouarde@bordeaux.inra.fr

Jean-Pierre L

Mark I

Sylvain D

ON THE ACCURACY OF LST MEASUREMENTS

1/15

Planetary Boundary

Layer

Surface Boundary

Layer

Pixel size

~x10m

~1km

convection Low frequency

(PBL) turbulence Scales: ~1 km, x min

High frequency (SBL) turbulence

Scales: 1m → x 10m, x secs

● LST is governed by surface energy budget and displays temporal

fluctuations generated by both SBL and PBL turbulence

● What do a instantaneous satellite measurement mean?

At high resolution ? for designing new missions in the TIR such as MISTIGRI, Hyspiri, THIRSTY…

At moderate resolution ? MODIS, VIIRS…

Land Surface Temperature and Atmospheric Turbulence

Mechanical interactions between the surface and atmosphere are well known….

Land Surface Temperature and Atmospheric Turbulence

4/16

…but they also exist in the thermal domain

Land Surface Temperature and Atmospheric Turbulence

• TIR camera A40-M FLIR

• 80° x 64.4° wide FOV

• Incl. ~60°/nadir

• NeDT 0.08 K at 30°C

• Acquisition: 12.5 Hz, 18mn

• 3D Young 81000V sonic anemometer

Bilos 44° 30' 03'‘N 0° 57' 20'‘W

SW France

Experimental

● Acquisition of high frequency time series at high spatial resolution from TIR cameras placed on masts or helicopter-borne

● Working with

brightness

● Geometric rectifications to superpose all images (with helicopter) → stacks of TIR images

● Reconstructing time series at different spatial resolution by agregation according to T⁴ scheme

● Analysis of temporal fluctuations according to resolution

2 approaches followed:

Modelling

● Use of a Large Eddy Simulation atmospheric model (ARPS) coupled with a SVAT model (MuSiCA)

● Equivalence between LST spatial variability within a domain and temporal variability at a given point within the domain

● Pine canopy

Analysis of LST fluctuations : methodology

Experimental set-up over agricultural crops

32 33 34 35 36 37 38 39

0 100 200 300 400 500 600 700 800 900 1000 1100 1200

Surface temperature (°C)

Time (s) 5 m 95 m

Spatial integration of small-scale fluctuations

→ The amplitude of the LST temporal fluctuations decreases with spatial resolution

Maize

37.5 38 38.5 39 39.5 40

4 6 8 10 12 14 16 18 20 22

0 1 2 3 4 5

4 6 8 10 12 14 16 18 20 22

Time (minutes from 15:55 UTC)

28 29 30 31 32 33 34

4 6 8 10 12 14 16 18 20 22

Time (minutes from 12:00 UTC) T air (°C)Horizontal wind (ms-1) Surface temperature (°C)

Maize (50m aggregation) Bare soil (50m aggregation)

Analysis of maximum cross correlation coefficient between LST and windspeed

→ prevailing effect of wind on LST temporal fluctuations No correlation with air temperature found

Experimental measurements (agricultural crops ): results

0 100 200 300 400 500m

INRA Le Bray site 44° 43' 01.50”N 0° 46' 09.00”W

Experimental measurements over pine stands

•Flux tower

•Gill R3 3D sonic anemometer

7/15

• Sept. 10, 1999

• Inframetrics M760

• 17 x 22° FOV

• NeDT < 0.2°C

• 2 acquisitions:

30mn, 10 Hz 2 directions (E, S)

x

x

20m

40m

Experimental measurements over pine stands

33 33.5 34 34.5 35 35.5 36

132 134 136 138 140 142 144 146 148 150 152 154 156 158 160 162 164

Ts (°C)

Time (minutes from 12:00 UTC)

34 34.5 35 35.5 36 36.5 37

168 170 172 174 176 178 180 182 184 186 188 190 192 194 196 198 200

Ts (°C)

Time (minutes from 12:00 UTC)

12:13-12:43 UTC

12:49-13:20UTC

±1.5°C

±1.5°C Resolution : ~20m

• August 16^th, 1995

• Inframetrics M760

• 80 x 60° FOV

• 700 m flight height

• Acquisition:

Measurements from helicopter over le Bray (pixel at tower site)

33.5 34 34.5 35 35.5 36

0 1 2 3 4 5 6 7 8 9 10 11 12

ST (°C)

Time (minutes after 12:50 UTC)

0 1 2 3 4 5 6 7 8

0 1 2 3 4 5 6 7 8 9 10 11 12

Horizontal windspeed (ms-1 )

Time (minutes after 12:50 UTC)

Resolution : ~60m

Cross correlation : ~ 6s lag ⇒~60m vith u ~3m^s-1

9/15

Experimental measurements over pine stands

● Equivalence between temporal fluctuations at a given location and spatial variability within the domain ● LES ARPS modem coupled with

MuSICA SVAT

● 3000x3000m x 2.5km LST

Vertical wind

LES simulation

● Simulation of a pine stand surface

● Spectral analysis

confirms the prevailaing effect of wind on LST fluctuations

MEA 5m MEA 95m

AD

MEA maximum error amplitude

DEP amplitude deviation from the mean temperature (absolute value)

Quantifying LST fluctuations and uncertainties

14/16

Key interpretation parameters

Statistics of the deviation

0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 DEP (°C)

0 1 2 3 4 5 6

0 20 40 60 80 100 120 140 160 180 200

Maximum Error Aamplitude (°C)

Spatial resolution (m)

maize bare soil Pine LES maize

bare soil

pine LES

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

Frequency (%)

Ts uncertainty (°C)

Seq 1 Seq 2 maize bare soil Bray 1995

Uncertainty :

< ±0.6 °C for ~70% measurements

< ±0.8 °C for ~85% measurements at 50m resolution

Lagouarde et al., RSE 2013

Lagouarde et al., RSE 2014 submitted

Quantifying LST fluctuations and uncertainties: experimental results

Decrease of amplitude of fluctuations for high spatial resolutions explained by integration effect of small size SBL eddies over the pixel

Larger PBL eddies contribute to LST uncertainty for resolutions > 50-100m

Results

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

0 50 100 150 200 250 300

Maximum error amplitude (MEA) (°C)

Spatial resolution (m)

Maximum MEA Minimum MEA Mean MEA Mean MEA - 1 std Mean MEA + 1 std

Quantifying LST fluctuations and uncertainties: LES simulation results

LES simulation confirms experimental results → decrease of MEA with spatial resolution

Impact of spatial resolution on DEP illustrated

0 10 20 30 40 50 60 70 80 90 100

Frequency (%) 7m

21m 35m 56m 105m 203m

Low resolution

high resolution

16/16

Results

● LST measurements are prone to uncertainties due to atmospheric turbulence.

● The uncertainty depends on spatial resolution : high resolution is dramatically affected by SBL turbulence

● At 50-60 m resolution, for measurements over vegetation : LST error > ± 0.6K for 20-25%

LST error > ± 0.8K for 10-15%

● Generalization to other surfaces and metorological conditions needed

● Case of water bodies to be investigated

Conclusions and recommendations

Recommandations

Care must be brought when using satellite data in combination with SVAT models (which generally have 30 mn to 1 hour time steps)

1

Uncertainty both on satellite data and ground measurements to be taken into account when performing validation exercises

2

Possibility of relaxing the NeDT specification (0.2 K for vegetation?) 4

Need of high revisit fo future TIR systems 5

3 Is it necessary to aim at very high resolution in the TIR (<50m) for continental biosphere sutdies?

For using satellite data…

For designing future TIR missions…

15/15