Retrievals of Land surface temperature
using LANDSAT8 and ASTER data over
Italian volcanic areas: comparison of
satellites and in situ measurements
M.F. Buongiorno, M. Silvestri, M. Musacchio
Istituto Nazionale di Geofisica e Vulcanologia
GlobTemperature User Consultation Meeting #4, Lisbon, 7th and 8th June 2016
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
Summary
•
•
•
•
•
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
General Approach
Temperature Emissivity Separation
(TES) Gillespie et al., 1998
(
Land Surface Temperature
LST
(
TERRA-ASTER
Landsat8
Barsi et al., 2003
Barsi, J.A., J.L. Barker, J.R. Schott. An Atmospheric Correction Parameter Calculator for a Single Thermal Band
Earth-Sensing Instrument. IGARSS03, 21-25 July 2003, Centre de Congres Pierre Baudis, Toulouse, France.
Gillespie, A. R., Matsunaga, T., Rokugawa,
S., & Hook, S. J.User
(1998).
Temperature and emissivity separation from
GLOBTEMPERATURE
Consultation
Advanced Spaceborne Thermal Emissionmeeting,
and Reflection
(ASTER) images. IEEE Transactions on
LisbonRadiometer
7-8 June 2016
Geoscience and Remote Sensing, 36, 1113-1125.
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Landsat 8 specific solution
Temperature Emissivity Separation
(TES) Gillespie et al., 1998
(
Land Surface Temperature
LST
(
TERRA-ASTER
Barsi et al., 2003
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
Landsat8
General approach
Test sites
Ground truth and validation
Data analysis
Barsi et al., 2003
τεLT + Lu + τ (1−ε)Ld = Ll
Where
τ is the atmospheric transmission;
ε is the emissivity of the surface;
LT is the radiance of a blackbody target of kinetic temperature T;
Lu is the upwelling or atmospheric path radiance;
Ld is the downwelling or sky radiance;
• GED ASTER Emissivity Map
• TERRA ASTER 05 standard product
• Ground measurement by means of FTIR
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
Conclusion
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
TERRA-ASTER Specific Solution
Temperature Emissivity Separation
(TES) Gillespie et al., 1998
(
Land Surface Temperature
LST
(
TERRA-ASTER
Barsi et al., 2003
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
Landsat8
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Temperature Emissivity Separation
• τεLT + Lu + τ (1−ε)Ld = Ll
By using TES (Gillespie et al, 1998), emissivity values is
obtained by the Maximum Minimum Difference (MMD)
algorithm where emissivities and temperature have been
calculated starting from e = 0.98 avoiding the
uncertainties due to radiance measured in n spectral
bands, there will be n+1 unknowns, n emissivity and one
surface temperature
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Test Sites: Mt Etna, Sicily (Italy)
Mt Etna is the largest active volcano in Europe with a diameter of 40x40 kmq and
elevation of abot 3350 m a.s.l. Towering above the city of Catania on the island of
Sicily, it has been growing for about 500,000 years.
Mt Etna has the longest period of documented eruptions in the world. Etna is noted
for the wide variety of eruption styles.
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Test Sites: Campi Flegrei, Campania (Italy)
The earliest known eruptive products are dated 47,000 years before present (BP). The Campi Flegrei caldera
formed following two large explosive eruptions, the massive Campanian ignimbrite about 36,000 years BP, and
the >40 cubic km Neapolitan Yellow Tuff (NYT) about 15,000 years BP. Following eruption of the NYT a large
number of eruptions have taken place from widely scattered subaerial and submarine vents. Most activity
occurred during three intervals: 15,000-9500, 8600-8200, and 4800-3800 years BP. Two eruptions have occurred
in historical time, one in 1158 at Solfatara and the other in 1538 that formed the Monte Nuovo cinder cone.
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Field Campaigns Instruments: emissivity
1,02
GROUND EMISSIVITY
1
0,99
Emissivity
Model 102F FTIR
(Fourier transform
infrared) spectrometer
1,01
0,98
0,97
0,96
0,95
0,94
0,93
8,00
9,00
10,00
11,00
12,00
13,00
14,00
Wavelenght
Campi Flegrei Field campaign: 28/09/2009
FTIR SPECIFICATIONS
Ground emissivity
0,98
Ground emissivity
Item
Parameter
Value
Units
Comments
0,97
0,97
2
Spectral
Range
2 - 16
Spectral
Resolution
(FWHH)
4
micromete Standard
rs
IR
wavenumb Standard,
ers
1 sec. scan
0,96
0,96
Emissivity
1
0,95
0,95
0,94
0,94
0,93
0,93
8,00
9,00
10,00
11,00
12,00
13,00
14,00
Wavelenght (micron)
Mt Etna
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
Field campaign: 07/10/2009
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Field Campaigns Instruments: Temperature
Thermoteknix System Ldt VisIR 640 thermal
camera, in the 8 -14µm spectral range.
The EVEREST 130,2L
is an infrared
thermometer in the
8 -14µm spectral
range.
Day time surface temperature variation collected by means of infrared
GLOBTEMPERATURE
User Consultation
portable thermometer.
Measurements
have been started at 10:00 UTC for
meeting,
Lisbon
7-8
June
2016
20 minutes during ASTER over passing (28/09/2009)
General approach
Test sites
Ground truth and validation
Data analysis
Day time
Day time Landsat8 - 27/10/2014 – 9.45
Night
time
Conclusion
h 12,40 Tatm = 22,9 U = 36%
Min
Max Mean
Landsat8
27/10/2014 (time
9.45 GMT)
19.521
27.936
22.798
ASTER
27/10/2014 (time
21.00 GMT)
17.731
27.047
20.289
ASTER - 27/10/2014 – 21.00
Night time
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Emissivity comparison: Solfatara
campaign (28/9/2009)
1,04
1,02
1
0,98
0,96
0,94
0,92
0,9
8
8,5
9
S7
S8
9,5
S19
S29
10
S31
Average
10,5
ASTER
EMISS_ASTER
11
ASTER-GED
11,5
12
TES
S7, S8, S19, S29, S31 and AVERAGE: Measurements collected with FTIR
Measurements collected with FTIR and convolved on ASTER SRF
ASTER-GED: emissivity retrieved by the ASTER-GED
EMISS_ASTER: emissivity retrieved by ASTER 05 data
Emissivity retrieved using TES algorithm
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
ASTER-GED emissivity map
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Temperature Cross Comparison
Solfatara field campaign 28/9/2009
Surface temperature from TES
Surface Temperature Map
Surface Kinetic Temperature from ASTER 08
GLOBTEMPERATURE User Consultation
meeting, Lisbon
7-8 June 2016 collected
Temperature
with Everest
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Cross Comparison: ASTER vs Landsat8
Night time
ASTER 18/07/2015
Night time
LANDSAT 19/07/2015
Night time
ASTER 20/07/2015
ROI on 56 points
Min
Max
Mean
Stdev
ASTER 18/07/2015
24.770935
36.675934
28.535006
3.011479
LANDSAT 8
19/07/2015
25.786102
35.495209
28.911593
2.598132
ASTER 20/07/2015
25.650024
37.518768
29.153285
2.816236
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Temperature Cross Comparison
Etna field campaign 07/10/2009
Validation with ground truth
SITE
LAT
LON
H
TIME
Piano delle Concazze 37.7653GLOBTEMPERATURE
15.0128 User 2793
10:46
Consultation
meeting, Lisbon 7-8 June 2016
T °C
T °C
STEMP
35.9±3.3
37.14
Test sites
Ground truth and validation
Data analysis
Conclusion
Temperature Cross Comparison
Etna field campaign 25/06/2012
Day time TES
General approach
Day time ASTER 08
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
General approach
Test sites
Ground truth and validation
Data analysis
Conclusion
Conclusions
• ASTER and L8 have been scaled at ASTER-TIR channel spatial resolution (90
m)
• Ground truth data are referred to smaller area (nominally a point)
• Using two different methods (TES, BARSI) the difference in retrieved
temperature is less than 3°C and 1°C respectively for Solfatara di Pozzuoli
and Mt Etna
• The results obtained are very useful to understand variability of LST
retrievals by remote sensing data
• They highlight the importance of precise emissivity inputs to the retrieval
procedures as well as the needs of systematic calibration areas for LST
satellite retrievals
• Ground data for Etna need still to be analyzed
• Comparison with GLOBTEMP data
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016
THANK YOU
GLOBTEMPERATURE User Consultation
meeting, Lisbon 7-8 June 2016