digital remote sensing - us forest service · • pixels in a raster format a pixel (picture...

Burn Severity and Remote Sensing

Digital Remote Sensing

BARC Use Training 2010

Overview

• Defining burn intensity and severity

• Measures of severity

• Remote sensing basics / Sensor properties

Monitoring Trends in Burn Severity Project, http://www.mtbs.gov

Fire Intensity

• The amount of energy or heat release per unit time or area and encompasses several specific types of fire intensity measures.

• Byram (1959): “The rate of energy or heat release per unit time, per unit length of fire front, regardless of its depth.” depth.”

Photo courtesy of NPS


Byram, G.M. 1959. Combustion of

forest fuels. In: Davis, K.P. (ed.).

Forest fire: control and use.

McGraw-Hill, New York. p. 61-89.

Fire (Burn) Severity

• The effect of a fire on ecosystem properties, often defined by the degree of mortality of vegetation.

– Relates to soil heating, large fuel and duff consumption, consumption of the litter and organic layer beneath trees and isolated shrubs, and mortality of buried plant parts.

• Degree to which a site has been altered or disrupted by fire; loosely, a product of fire intensity and residence fire; loosely, a product of fire intensity and residence time.

Photo courtesy of Stefan Doerr


Soil Burn Severity

• The fire-induced changes in physical, chemical, and biological soil properties that impact hydrological and biological soil functions

Photo courtesy of Stefan Doerr


Example in Pictures


Field Perspective

Ground-based severity assessments may include:

• Composite Burn Index (CBI)

• Hiking through and observing burn scar mosaic

• Water repellency tests


Satellite Perspective

Imagery Severity


Connecting the Dots

• How do we connect pixels in a satellite image to burn severity on the ground?


What is Remote Sensing?

Remote Sensing can be defined as: the collection and interpretation of information about objects based on the measurement of electromagnetic energy reflected or emitted from those objects.

We can collect remotely sensed data in a number of ways: Our eyes are sensitive to a portion of the EM spectrum, airborne and spaceborne sensors can carry instruments to record EM energy...


What is EM Energy?

0.4 0.5 0.6 0.7 µm

VisibleVisible λλλλλλλλ

Wavelength (µm) Wavelength (µm)

X-Rays Ultraviolet TV/RadioMicrowaveInfrared

EM energy is a continuum which we (somewhat arbitrarily) classify according to wavelength. Wavelengths extend from very, very short (cosmic and X rays) to very, very long (thermal, radar, etc...).

10-6 10-110-5 10-4 10-3 10-2 1 10 102 103 104 105 106 107 108


Remote Sensing and EM Energy

Conifer

Asphalt

visually distinguish one thing from another.

Remote sensing relies on the fact that different targets have unique responses to EM energy—allowing us to

Water


Response to EM Energy

Spectral Response Curves, aka Spectral Signatures

Graphically, the spectral reflectance of green vegetation in the visible wavelengths may be the visible wavelengths may be represented as shown

0.4

Wavelength (µm)

2.6

Reflectance


The Sun Emits a Full Spectrum of EM Energy

Thus our tree has a spectral signature that extends beyond the visible

X-Rays Ultraviolet TV/RadioMicrowaveInfrared

Wavelength (µm) Wavelength (µm)

10-6 10-110-5 10-4 10-3 10-2 1 10 102 103 104 105 106 107 108


Response to EM Energy

Spectral response curve of typical vegetation from 0.4 to 2.6 µm

Relatively high green response due to

High near infrared response due to healthy plant cell structure

Relatively low responses in the mid-infrared due to water

absorption

0.4

Wavelength (µm)

2.6

Reflectance

response due to chlorophyll pigmentation


Typical Spectral Signatures

Typical Spectral Response Curves in the 0.4 to 2.6 µm Region...

Healthy Vegetation

Dry, Bare Soil

0.4

Wavelength (µm)

2.6

Reflectance Clear Water


Healthy Vegetation vs. Burned Areas

Exploiting Spectral Response Curves

High Burn SeverityUnburned

Mod. Burn Severity

Low Burn Severity

The goal of remote sensing is to take advantage of differences in spectral response curves to distinguish one thing from another.

0.4

Wavelength (µm)

2.6

Reflectance


Important Satellite Sensor Properties

• Spatial Properties

– Resolution• How small of an object can we see?

– Extent• How large of an area is covered?

• Revisit time• How often can we see the same area?• How often can we see the same area?

• Spectral sensitivity• How many “colors” can we see?


Sensor Spatial Properties

• Spatial Resolution

– Measured as the Ground Sample Distance or more commonly Pixel Size• The distance on the ground covered by the Instantaneous Field Of

View (IFOV) of the sensor detectors



• Pixels in a Raster Format

A Pixel (picture element) is an individual cell in a raster image.

Columns Single Pixel

individual cell in a raster image. Each pixel has three dimensions:1. Length2. Width3. Digital number.The value of the digital number relates directly to the average integrated brightness of all the surface objects and material contained within the pixel.

Rows


Sensor Properties—Spatial Resolution

• Spatial Resolution - Pixel Sizes of selected sensors

Landsat (30 m)

SPOT 1-4 (20 m)

30m30m 20m20m 10m10m

SPOT 5 (10 m)

Ikonos (4 m)

Quickbird (2.4 m)



• Spatial Extent (Area Covered)

– Generally, there is a direct relationship betweenpixel size and image extent• Sensors that have a large IFOV (large pixel size) usually

produce imagery that covers large areas

• Sensors that have a small IFOV (small pixel size) usually produce imagery that covers small areas

Sensor Pixel Size (m) Extent (sq km)

AWiFS 56 547,600

Landsat 30 34,225

SPOT 5 10 3,600

Quickbird 2.4 272


Sensor Properties—Revisit Time

• How often does the sensor gather an image of the same ground area? Depends on:

– Orbital characteristics

– Image Swath width

– Off-nadir viewing capabilities (i.e., pointable optics)

– Number of satellites in the family


Sensor: Landsat/ASTER SPOT 4,5 AWiFS Quickbird/IKONOS

Revisit Time 8 days 3-4 days 5 days 3-4 days

Notes:

Landsat 7 and ASTER

have a revisit time of 16

days each.

Landsat 5 images an

area 8 days after

Landsat 7.

Imagery is typically

acquired 48-72 hours

after an order is

submitted. Clouds

and smoke can delay

useful acquisition.

Imagery is typically

acquired 48-72 hours

after an order is

submitted. Clouds

and smoke can delay

useful acquisition.

Areas can be

revisited every 2 to

11 days depending on

latitude and look angle

tolerance.

Sensor Properties—Spectral Sensitivity

• Spectral Sensitivity:

– The size, number, and position of imaging bands.

– How many “colors” the sensor sees

– Example:

Example:

.1 .4 .5 .6 .7 .8 .9 1 1.5 2 3 4 5 6 7 8 9 10 11 12 µµµµ

UV Near Infrared SWIR Far InfraredMid Infrared

Sensor 1

Sensor 2

Example:

Relatively Fine Spectral Resolution

Relatively Coarse Spectral Resolution



• Multispectral Imagery in Ecosystem Management

.1 .4 .5 .6 .7 .8 .9 1 1.5 2 3 4 5 6 7 8 9 10 11 12 µµµµ

UV Near Infrared SWIR Far InfraredMid Infrared

Longer (e.g., Radar) wavelengths: Surface Texture, Interferometry, topography

Visible Region (Blue, Green, Red): Cultural features, soil versus water, hydrography, vegetation.

Near Infrared (Reflected Infrared): Vegetation discrimination, biomass, soil,

snow from clouds

Shortwave-Infrared (Partly reflected-Partly emitted): Moisture absorption, the high temperature thermal window, wildfires, vehicles, exhausts.

Far Infrared: Includes the longwave thermal window, vegetation stress, thermal



• Spectral Sensitivity of Common Sensor Systems

SPOT 4 (20m)

SPOT 5 (10m)

.1 .4 .5 .6 .7 .8 .9 1 1.5 2 3 4 5 6 7 8 9 10 11 12 µµµµm

UV Near Infrared Mid IR Far InfraredSWIR

Moderate Resolution

Landsat (30m)

SPOT 5 (10m)

Ikonos (4m)

Quickbird (2.4m)

AWiFS (56m)

ASTER (30m)

Moderate Resolution

High Resolution


Healthy Vegetation vs Burned Areas

• Exploiting Spectral Response Curves

Healthy VegetationBurned Areas

0.4

Wavelength (µm)

2.6

Reflectance

Key spectral differences to exploit. What is the best way to take advantage of these differences ? --i.e., is there a way to accentuate the differences?

Landsatband 4

Landsatband 7


Band Ratios used for Severity Mapping

• Normalized Burn Ratio (NBR)

(B4 – B7) / (B4 + B7)

Pre Refl

Pre NBR

Post Refl

Post NBR


Change Detection

• Differenced Normalized Burn Ratio (dNBR)

Prefire NBR – Postfire NBR


Questions??

Exercise 1: Image Viewing Tools and Techniques

ExerciseExercise

digital remote sensing - us forest service · • pixels in a raster format a pixel (picture...

Documents