Lecture 3: Thermal remote sensing

Question 1

Types of forest fires

Answer 1

3 types of fire:
➢ Deforestation-related (preparation for agriculture)
➢ Slash-and-burn (agricultural burns)
➢ Uncontrolled fires reaching the forest
→ CAUSES = deforestation + agriculture
Role of climate change:
➢ Towards longer dry seasons

Question 2

Thermal infrared (TIR) radiation

Answer 2

Theoretical background
➢ All materials at temperature
above absolute 0 (0 K or -273.15°
C) continuously emit EMR.
➢ The Earth with its ambient
temperature of ~300 K has its
peak energy emission in the TIR
region at 9.7 µm.
➢ 2 atmospheric windows: 3-5 µm
and 8-14 µm

Question 3

Planck’s Blackbody Radiation Law

Answer 3

= EMR emitted from a blackbody at
a certain wavelength, as a function
of its absolute temperature
→ If the EMR from an object can be
measured at a given wavelength, we
can theoretically retrieve the
temperature of that object

Question 4

Stefan-Boltzmann Law

Answer 4

= Total EMR emitted from a
blackbody as a function of its
absolute temperature
= area under the Plank’s curve
→ The higher the temperature of the
radiator, the greater the total amount
of radiation it emits

Btot = 𝜎 T^4
Btot = total radiant exitance (W m-2)
𝜎 = Stefan-Boltzmann constant
(5.6697 x 10-8 W m-2 K-4)
T = absolute temperature (K)

Question 5

Theory vs. reality:
In real life, nothing is a blackbody

Answer 5

More “graybodies” with an emissivity ratio < 1
ε = Mr/Mb < 1
ε = emissivity
Mr = radiance emitted by a radiative body
Mb = radiance emitted by a blackbody at the same temperature

It is even more complex: Emissivity can evolve with wavelength

Question 6

Typical emissivity values

Answer 6

Emissivity depends on:
▪ Colour
▪ Surface roughness
▪ Moisture content
▪ Compaction
▪ Field of view

Question 7

Atmospheric effects

Answer 7

Atmosphere between ground surface
and TIR sensor → it can modify the
apparent measured level of radiations
by:
➢ Absorption (→ appears colder)
➢ Scattering (→ appears colder)
➢ Emission (→ multi-source radiation)
Absorption and scattering are affected by:
➢ Atmospheric path length
➢ Regional+local meteorological conditions
➢ Specificities of the surface
➢ Altitude

Question 8

Wien’s Displacement Law

Answer 8

= Wavelength at which the
maximum spectral radiant exitance
occurs
λmax = A / T
A = Wien’s constant (2897.8 µm K)
→ The higher the temperature of the
radiator, the more λmax towards
shorter wavelengths

Question 9

Principle of hotspot detection

Answer 9

→ Taking advantage of Wien’s Displacement Law
GROUND SURFACE: B4µm < B12µm
FIRE / HOT LAVA: B4µm > B12µm

➢ Use of an index with a threshold to
detect hotspots
Example of the Normalized Thermal
Index (NTI)
NTI = (𝐿4 −𝐿12)/(𝐿4+𝐿12)
➢ The NTI threshold can be either fixed or dynamic

Question 10

Radiant heat

Answer 10

= Transforming spectral radiance into radiative power
e.g., Wooster et al. (2003) for MODIS
RP = 1.89 x 107 (L4 – L4,bg)
Unit = J s-1 or W
➢ For fires, we talk about Fire Radiative Power (FRP)
➢ For volcanoes, we talk about Volcanic Radiative Power (VRP)

Question 11

Sentinel-3 SLSTR imagery

Answer 11

✓ Sun-synchronous
✓ Systematic global acquisition
✓ Daily revisit time
Instruments:
▪ OLCI: Ocean and Land Colour instrument
▪ SLSTR: Sea and Land Surface Temperature Radiometer
▪ SRAL: SAR Radar Altimeter
▪ MWR: Microwave radiometer
Main missions:
▪ Ocean and land surface colour (→ OLCI)
▪ Sea and land surface temperature (→ SLSTR)
▪ Sea surface topography (SRAL + MWR)

Question 12

Landsat-type satellites

Answer 12

Landsat-type satellites (e.g., Landsat-8, Sentinel-2)
➢ Moderate spatial resolution (20-30m)
➢ Combined: high temporal resolution (2-5 days)
➢ Hotspot can be detected in SWIR range as well
Advantages:
▪ Spatial resolution → interpretation of heat source
▪ Data available on cloud platforms (e.g., GEE)
Limitations:
▪ Few nighttime acquisitions
▪ Daytime: contamination by reflectance
▪ Sentinel-2: no TIR band

Question 13

SWIR hotspot detection: HOTMAP

Answer 13

HOTMAP
Purpose: Detect hot pixel clusters using shortwave infrared (SWIR) bands.
▪ Identify pixels satisfying α and β.
▪ Cluster them.
▪ Retain only clusters where at least one pixel passes α.
Interpretation: Understand red (true hits), blue (false alarms), and green boxes (clusters).

Question 14

SWIR hotspot detection: NHI

Answer 14

NHI (Normalized Hotspot Index)
Purpose: Detect SWIR hotspots and correct for atmospheric or sensor errors.
▪ compares radiance values in different spectral bands
▪ measuring how much hotter one band is compared to another, normalized by their total energy

Two types:
NHI -SWIR
→ Detect hot pixels with high thermal emission in 2.2 µm compared to 1.6 µm.
→ Higher NHI = stronger hotspot signal.
NHI -SWNIR
→ Secondary check: helps detect false positives due to vegetation or terrain reflection.
→ Filters out non-hot surfaces that just reflect strongly in the SWIR.