Landsat 8 LST Calculation Tool

Calculate Land Surface Temperature (LST) from Landsat 8 thermal infrared data. Understand the essential inputs and intermediate steps for accurate remote sensing analysis.

Landsat 8 LST Calculator

Intermediate Values Table

Parameter	Value	Unit	Description
Digital Number (DN)	—	–	Raw pixel value from thermal band.
Radiometric Calibration K1	—	W/(m²·sr·µm)	Planck’s Law constant K1.
Radiometric Calibration K2	—	K	Planck’s Law constant K2.
Band Wavelength (λ)	—	µm	Effective wavelength of the thermal band.
Radiant Spectral Luminance (Lλ)	—	W/(m²·sr·µm)	Energy emitted per unit area, per unit solid angle, per unit wavelength.
Radiometric Temperature (TB)	—	K	Temperature derived directly from radiance, assuming blackbody.
Surface Emissivity (ε)	—	–	Ratio of emitted radiance to that of a blackbody at the same temperature.
Calculated LST	—	°C	Corrected temperature accounting for emissivity.

What is Landsat 8 LST Calculation?

The Landsat 8 Land Surface Temperature (LST) calculation is a crucial process in remote sensing that estimates the thermal radiance emitted from the Earth’s surface, allowing scientists to derive the actual temperature of that surface. Landsat 8, operated by the USGS, carries the Thermal Infrared Sensor (TIRS) which captures data in specific thermal infrared wavelengths. Unlike air temperature measured by weather stations, LST represents the radiative temperature of the ground, which is influenced by factors like solar radiation, surface properties, and thermal inertia.

This calculation is vital for a wide range of applications. Environmental scientists use LST data to monitor urban heat islands, track drought conditions, study wildfire behavior, and analyze climate change impacts. Agricultural experts utilize it for irrigation management and crop stress assessment. Hydrologists employ LST to understand evaporation rates and water body temperatures. Geologists might use it for volcanic activity monitoring. Essentially, anyone studying surface energy balance, thermal pollution, or the physical state of the Earth’s surface can benefit from accurate LST estimations.

A common misconception is that LST is the same as atmospheric or air temperature. While related, they are distinct. Air temperature is measured a few meters above the ground, influenced by air circulation, while LST is the temperature of the surface itself, often much hotter or colder than the air temperature, especially under direct sunlight or during radiative cooling at night. Another misconception is that the raw digital numbers (DNs) from the thermal band directly represent temperature; this is incorrect, as they must be converted through radiometric calibration and further processed using emissivity values.

Landsat 8 LST Calculation Formula and Mathematical Explanation

The estimation of Land Surface Temperature (LST) from Landsat 8 thermal infrared data involves a multi-step process. The primary goal is to convert raw Digital Numbers (DNs) recorded by the sensor into physically meaningful radiance values, then into brightness temperature, and finally correct for the surface’s emissivity to obtain the true Land Surface Temperature.

Step 1: Radiometric Calibration to Radiance

The raw DN values recorded by the Landsat 8 TIRS sensor are not directly usable for temperature calculations. They must first be converted to spectral radiance (Lλ). This conversion uses pre-launch calibration coefficients, typically found in the image metadata file (MTL file). The formula is:

Lλ = ( (LMAX - LMIN) / (QMAX - QMIN) ) * (Q_i - QMIN) + LMIN

Where:

• Lλ is the spectral radiance at the sensor’s aperture (W/(m²·sr·µm)).
• LMAX is the spectral radiance corresponding to QMAX (W/(m²·sr·µm)).
• LMIN is the spectral radiance corresponding to QMIN (W/(m²·sr·µm)).
• QMAX is the maximum quantized calibrated digital number (usually 255 for 8-bit data, but check metadata).
• QMIN is the minimum quantized calibrated digital number (usually 1 for 8-bit data, but check metadata).
• Q_i is the DN value of the pixel (the input value).

Note: For Landsat 8, the constants K1 and K2 are often provided in the metadata, which are used in the next step to directly convert DN to Brightness Temperature, bypassing the intermediate radiance calculation if only temperature is needed. However, understanding radiance is fundamental.

Step 2: Conversion to Brightness Temperature (T_B)

Using the calculated spectral radiance (Lλ) and the sensor-specific calibration constants K1 and K2 (also found in the metadata, specific to each thermal band like B10 or B11), we can calculate the Brightness Temperature (T_B). This is the temperature of a blackbody that would emit the same radiance measured by the sensor. The formula is derived from Planck’s Law:

TB = K2 / ln( (K1 / Lλ) + 1 )

Where:

• TB is the Brightness Temperature (Kelvin, K).
• K1 and K2 are the effective radiometric calibration constants (W/(m²·sr·µm) and K, respectively) for the specific thermal band.
• Lλ is the spectral radiance calculated in Step 1.
• ln is the natural logarithm.

*Direct DN to T_B Conversion (Commonly Used for Landsat 8)*:
Many resources and software use a direct conversion from DN to Brightness Temperature using K1 and K2 constants. The formula often presented is:
TB = K2 / (ln((K1 / (( (LMAX - LMIN) / (QMAX - QMIN) ) * (DN - QMIN) + LMIN)) + 1))
However, if the software or tool directly provides K1 and K2 for temperature conversion (e.g., `REFLECTANCE_MULT_BAND_x`, `REFLECTANCE_ADD_BAND_x` for radiance, and `K1_CONSTANT_BAND_x`, `K2_CONSTANT_BAND_x` for temperature), you can simplify the calculation. The calculator uses a simplified form that assumes you’ve already obtained the correct Lλ, or directly applies K1 and K2 if your input is DN.

Step 3: Land Surface Temperature (LST) Correction using Emissivity

Brightness Temperature (T_B) is not the true Land Surface Temperature because most natural surfaces are not perfect blackbodies; they have emissivity (ε) less than 1. Emissivity is the ratio of the energy radiated by a surface to the energy radiated by a perfect blackbody at the same temperature. To get the true LST, T_B must be corrected for emissivity. A common approach, particularly for single-channel methods like this, is the Emissivity Normalization Method:

LST = TB / ( 1 + ( (λ * TB) / ρ ) * ln(ε) )

Where:

• LST is the Land Surface Temperature (Kelvin, K).
• TB is the Brightness Temperature (K) from Step 2.
• λ is the effective wavelength of the thermal band (µm). For Landsat 8 TIRS Band 10, λ ≈ 10.9 µm; for Band 11, λ ≈ 12.0 µm.
• ε is the land surface emissivity (dimensionless, typically 0.95-0.99).
• ρ (rho) is a constant derived from Planck’s Law: ρ = hc/σ, where h is Planck’s constant (6.626 x 10^-34 J·s), c is the speed of light (2.998 x 10⁸ m/s), and σ is the Boltzmann constant (1.382 x 10^-23 J/K). The value of ρ is approximately 14380 µm·K.
• ln is the natural logarithm.

The final result is usually converted from Kelvin to Celsius by subtracting 273.15.

Variables Table

Landsat 8 LST Calculation Variables

Variable	Meaning	Unit	Typical Range / Value
DN (Qi)	Digital Number	–	0 – 65535 (for 16-bit) or 0-255 (for 8-bit scenes)
K1, K2	Radiometric Calibration Constants	K1: W/(m²·sr·µm) K2: K	Band 10: K1=774.8853, K2=1321.0789 Band 11: K1=480.8883, K2=1261.1832
Lλ	Spectral Radiance	W/(m²·sr·µm)	Depends on scene, typically 0 to a few W/(m²·sr·µm)
TB	Brightness Temperature	K / °C	Typically 260 K to 350 K (approx. -13°C to 77°C)
ε	Surface Emissivity	–	0.90 – 0.99 (land surfaces)
λ	Effective Wavelength	µm	~10.9 µm (Band 10), ~12.0 µm (Band 11)
ρ	Constant (hc/σ)	µm·K	~14380
LST	Land Surface Temperature	K / °C	Depends on location and time, can range widely

Practical Examples (Real-World Use Cases)

Example 1: Urban Heat Island Monitoring

Scenario: Analyzing thermal pollution in a major city during a summer heatwave using a Landsat 8 image. A specific pixel over a dense commercial area shows a DN value of 225 in Band 10. The surface is primarily asphalt and concrete, with an estimated emissivity of 0.96.

Inputs:

• Thermal Band: B10
• DN Value (Q_i): 225
• Emissivity (ε): 0.96
• K1 (for B10): 774.8853
• K2 (for B10): 1321.0789
• Band Wavelength (λ for B10): 10.9 µm
• ρ: 14380 µm·K

Calculation Steps (using the calculator):

1. Radiance (Lλ) calculation (intermediate).
2. Brightness Temperature (T_B) calculation: T_B ≈ 305.5 K (≈ 32.3 °C).
3. LST calculation: LST = 305.5 K / (1 + ( (10.9 µm * 305.5 K) / 14380 µm·K ) * ln(0.96) ) ≈ 305.5 K / (1 + 0.00224 * -0.0408) ≈ 305.5 K / 0.9999 ≈ 305.5 K
4. Convert to Celsius: LST ≈ 305.5 – 273.15 = 32.35 °C

Result Interpretation: The calculated LST of 32.35 °C indicates a significantly high surface temperature for the commercial area, likely due to the absorption properties of asphalt and concrete, minimal vegetation cover, and trapped heat within urban structures. This value is substantially higher than the concurrent air temperature, highlighting the urban heat island effect. Comparing this LST to pixels over parks or water bodies (with higher emissivity and possibly more shade) would quantify the intensity of the UHI at this location.

Example 2: Agricultural Drought Monitoring

Scenario: Assessing water stress in a farmland region using Landsat 8 imagery during a dry season. A pixel over a cornfield shows a DN value of 180 in Band 10. Healthy vegetation typically has a high emissivity, estimated here at 0.97.

Inputs:

• Thermal Band: B10
• DN Value (Q_i): 180
• Emissivity (ε): 0.97
• K1 (for B10): 774.8853
• K2 (for B10): 1321.0789
• Band Wavelength (λ for B10): 10.9 µm
• ρ: 14380 µm·K

Calculation Steps (using the calculator):

1. Radiance (Lλ) calculation (intermediate).
2. Brightness Temperature (T_B) calculation: T_B ≈ 297.2 K (≈ 24.0 °C).
3. LST calculation: LST = 297.2 K / (1 + ( (10.9 µm * 297.2 K) / 14380 µm·K ) * ln(0.97) ) ≈ 297.2 K / (1 + 0.00225 * -0.03045) ≈ 297.2 K / 0.99993 ≈ 297.2 K
4. Convert to Celsius: LST ≈ 297.2 – 273.15 = 24.05 °C

Result Interpretation: The calculated LST of 24.05 °C suggests that the cornfield is relatively cool. This could indicate sufficient soil moisture, as transpiring plants tend to keep the surface cooler. If this LST were significantly higher than surrounding vegetated areas or historical averages for this period, it might suggest water stress, where the lack of transpiration leads to increased surface temperature. Comparing LST across different fields or over time helps identify areas needing attention for irrigation or management.

How to Use This Landsat 8 LST Calculator

Our Landsat 8 LST Calculator simplifies the complex process of deriving Land Surface Temperature from satellite imagery. Follow these steps for accurate results:

1. Gather Landsat 8 Data: Ensure you have a Landsat 8 scene and know the specific thermal infrared band you are using (typically Band 10 or Band 11).
2. Find Metadata: Access the Landsat 8 metadata file (MTL file) for your scene. This file contains the essential calibration constants (K1 and K2) specific to the thermal band.
3. Input Thermal Band Name: Enter the name of the thermal band (e.g., ‘B10’ or ‘B11’) into the ‘Thermal Band Name’ field.
4. Enter Calibration Constants: Input the K1 and K2 values corresponding to your chosen thermal band from the metadata file into the respective fields.
5. Provide DN Value: Identify a pixel of interest in your thermal band image and enter its Digital Number (DN) value. You can obtain this value using image processing software like QGIS or ENVI.
6. Estimate Emissivity: Determine the surface emissivity (ε) for your pixel’s land cover type. This is a critical input. Typical values range from 0.95 to 0.99 for vegetation and soil, but can be lower for water (around 0.99) or higher for specific materials. Consult remote sensing literature or land cover maps for appropriate values.
7. Click Calculate: Press the ‘Calculate LST’ button.

Reading the Results:

• Main Result (LST): The primary output is the estimated Land Surface Temperature in degrees Celsius (°C).
• Intermediate Values: Radiance (Lλ), Brightness Temperature (T_B), and the effective Emissivity used are displayed, showing the key steps in the calculation.
• Table Data: A detailed table breaks down all input parameters and intermediate results for clarity and verification.
• Chart: The dynamic chart visually represents how the Brightness Temperature and Emissivity influence the final LST.

Decision-Making Guidance: The calculated LST can be used to:

• Identify hot spots (e.g., urban heat islands, wildfire fuel).
• Assess vegetation health and water stress (cooler LST often means healthier, transpiring vegetation).
• Monitor drought conditions.
• Analyze thermal pollution in water bodies.
• Improve land surface modeling and climate studies.

Always compare your LST results with air temperature data, other LST products, or ground-truth measurements for comprehensive analysis.

Key Factors That Affect Landsat 8 LST Results

Several factors can significantly influence the accuracy and interpretation of Land Surface Temperature (LST) derived from Landsat 8:

1. Surface Emissivity (ε): This is arguably the most critical factor after the raw sensor calibration. Different surface types (water, soil, vegetation, concrete, asphalt) have varying emissivities. Inaccurate emissivity estimation will directly lead to errors in LST. For complex land cover within a pixel, spectral mixture analysis might be needed for more precise emissivity determination.
2. Atmospheric Effects: The atmosphere absorbs and emits thermal radiation, affecting the signal reaching the satellite. Water vapor, clouds, and aerosols can cause significant temperature biases. While single-channel methods like the one used here are simpler, they often rely on atmospheric correction models or average atmospheric profiles, which might not perfectly match the specific acquisition time and location. Multi-channel methods (using multiple thermal bands) or radiative transfer models can provide more accurate atmospheric correction.
3. Sensor Calibration Accuracy: The accuracy of the K1 and K2 constants provided in the metadata is paramount. Any drift or error in the sensor’s radiometric calibration directly translates to errors in radiance and subsequently brightness temperature, impacting the final LST. Regular calibration and validation by agencies like USGS are essential.
4. Effective Wavelength (λ) and ρ: The choice of the effective wavelength for the thermal band (λ) and the constant ρ influence the LST correction step. Using the correct values for the specific band (B10 vs. B11) is crucial. Small variations might occur based on different studies defining these values, but the standard values are generally accepted.
5. Time of Acquisition: Landsat 8 has a fixed overpass time (around 10:00 AM local time). LST varies dramatically throughout the day. Results from this time represent the surface temperature during mid-morning, which might differ significantly from daytime highs or nighttime lows. For comprehensive analysis, LST data from different times or sensors with diurnal coverage might be needed.
6. Pixel Heterogeneity: A single Landsat 8 pixel (30m resolution for reflective bands, 100m resampled to 30m for thermal) can contain a mix of surface types (e.g., buildings, roads, grass, soil). The derived LST represents an average temperature for that pixel, which may not accurately reflect the temperature of a specific sub-pixel feature. Higher spatial resolution thermal data might be required for detailed urban or landscape analysis.
7. Topographic Effects: In areas with significant elevation changes, slope and aspect can influence the amount of solar radiation received by the surface, affecting its temperature. These effects are typically not accounted for in basic single-channel LST algorithms but can be important in mountainous regions.

Frequently Asked Questions (FAQ)

What is the difference between Brightness Temperature (T_B) and Land Surface Temperature (LST)?

Brightness Temperature (T_B) is the temperature calculated directly from the thermal radiance measured by the satellite sensor, assuming the surface is a perfect blackbody (emissivity = 1). Land Surface Temperature (LST) is the actual physical temperature of the surface, which requires correcting T_B for the surface’s non-blackbody emissivity. LST is generally lower than T_B for surfaces with emissivity less than 1.

How do I find the K1 and K2 constants for Landsat 8?

The K1 and K2 constants are specific to each thermal band (Band 10 and Band 11) and are provided in the Landsat 8 metadata file (MTL file) for each scene. Look for parameters like `K1_CONSTANT_BAND_10`, `K2_CONSTANT_BAND_10`, `K1_CONSTANT_BAND_11`, and `K2_CONSTANT_BAND_11`.

What is a typical range for Land Surface Emissivity (ε)?

For land surfaces, emissivity typically ranges from 0.90 to 0.99. Vegetation and soil generally have high emissivities (0.95-0.99). Water has a very high emissivity, close to 0.99. Barren surfaces like rock or asphalt might have slightly lower values, but generally remain above 0.90. Accurate emissivity estimation depends heavily on land cover classification.

Can I use this calculator for Landsat 7 or Sentinel satellites?

This calculator is specifically designed for Landsat 8 using its standard thermal bands (B10/B11) and associated constants. Landsat 7 (ETM+), Sentinel-2 (no thermal band), or Sentinel-3 (different thermal bands and calibration) require different constants, algorithms, and potentially different input parameters. You would need a specialized calculator for those sensors.

What are the limitations of single-channel LST algorithms?

Single-channel methods, like the one used here, are simpler but have limitations. They are more sensitive to errors in atmospheric correction and emissivity estimation. Multi-channel methods, using two thermal bands, can better account for atmospheric effects and potentially derive both emissivity and LST simultaneously, leading to higher accuracy, especially under varying atmospheric conditions.

Does the time of day affect the LST result?

Yes, significantly. Landsat 8 acquires data around mid-morning. Surface temperature fluctuates greatly throughout the day due to solar heating and cooling. The LST value you calculate represents this mid-morning temperature, which will differ from peak afternoon temperatures or nighttime temperatures.

How does cloud cover impact LST calculations?

Clouds completely block thermal infrared radiation from reaching the satellite, rendering LST calculations impossible for cloudy pixels. Even thin cirrus clouds can significantly affect the accuracy of LST estimates in adjacent clear pixels due to scattering and absorption. Cloud masking is a critical pre-processing step.

Is it better to use Band 10 or Band 11 for LST calculation?

Both bands can be used, but Band 10 (approx. 10.9 µm) is generally preferred for LST estimation because it experiences less atmospheric influence (particularly from water vapor) compared to Band 11 (approx. 12.0 µm). However, using both bands in a split-window algorithm can improve accuracy by better characterizing atmospheric effects.

What are the units of the final LST result?

The calculator provides the final Land Surface Temperature in degrees Celsius (°C). The intermediate Brightness Temperature is calculated in Kelvin (K), which is then converted to Celsius.

Related Tools and Internal Resources

• Vegetation Index Calculator

  Calculate various vegetation indices like NDVI and EVI using Landsat imagery to assess vegetation health and density.

• Surface Albedo Calculator

  Estimate the surface albedo, which measures how much solar radiation is reflected by a surface, using Landsat data.

• Beginner’s Guide to Remote Sensing

  An introductory overview of remote sensing principles, data sources, and common applications.

• Comparison of GIS Software

  Explore popular Geographic Information System software options for analyzing satellite imagery and spatial data.

• Landsat Data Processing Workflow

  A detailed guide on the steps involved in pre-processing Landsat imagery for analysis.

• Urban Heat Island Analysis with Remote Sensing

  Learn how LST and other remote sensing techniques are used to study and mitigate urban heat islands.