test

hist-match

test

1/2

or the analysis of imagery taken by satellites and planes

is an effective tool for measuring the path of tornado damage and assessing damage to infrastructure and property.

Remote sensing

Classification

or the assignment of different types of land cover to imagery encompasses several techniques which can identify a tornado path.

High-Resolution Orthographic Imagery from MassGIS

0.3 meter spatial resolution
Recorded 2012 - 2013 from aircraft
4 Spectal Bands: Near-Infrared, Red, Green, Blue
183 individual grids downloaded and decompressed to cover rectangular area of the June 1st tornado

Data Aquisition

Image Processing

Using Hexagon Geospatial's ERDAS IMAGINE software:

Imported individual JPEG2000 files to *.img

Combined tiles using Mosaic Pro and histogram matching color correction

Grids were recorded at different times, resulting in color imbalance. Color correction was needed for consistent image analysis.

JUNE 1ST, 2011

TORNADO

A 2015 STUDY IN

REMOTE SENSING

SPRINGFIELD

MONSON

STURBRIDGE

February 2015:

The author captures the snow-covered path of destruction from above Sturbridge, MA

Image Processing

mosaicpro

color-correct

hist-match

mosaicpro

1/3

Resulting in:

Several techniques can be used to better identify the path of damage:

False color RGB (near-infrared, red, green)

In the imagery's native resolution, individual trees which have been downed by the storm are clearly visible.

Image Analysis

Subset to conventional shape (normal color RGB):

A standard Normalized Difference Vegetation Index (NDVI) produces a panchromatic image representing the ratio of red to near-infrared wavelengths reflected by vegetation. A darker area, indicating less vegetation, is still visible on the damage path after four years.

Image Analysis

A supervised classification with nine training signatures based on the above false color RGB produces a less clear result. Since the imagery was taken in early spring, it is difficult to differentiate between living deciduous stands and blowdown from the tornado.

However, isolating the layer identifying conifer stands reveals a distinct lack of this vegetation along the damage path.

...As does isolating the layer identifying bare earth or exposed pavement.

Conclusions

Of the three methods used above, the NDVI appears to produce the most distinctive results.
In addition to NDVI analysis, academic studies have also had good results using Primary Component Analysis (PCA) and Object-Oriented Approach to identify tornado tracks using before and after imagery.
Another application of remote sensing not addressed in this case study, satellite imagery and aerial photos are used by governments and the insurance industry to assess property damage after major storm events, such as a tornado.

or the analysis of imagery taken by satellites and planes

is an effective tool for measuring the path of tornado damage and assessing damage to infrastructure and property.

Remote sensing

Classification

or the assignment of different types of land cover to imagery encompasses several techniques which can identify a tornado path.

High-Resolution Orthographic Imagery from MassGIS

0.3 meter spatial resolution

Recorded 2012 - 2013 from aircraft

4 Spectal Bands: Near-Infrared, Red, Green, Blue

183 individual grids downloaded and decompressed to cover rectangular area of the June 1st tornado

Data Aquisition

Image Processing

Using Hexagon Geospatial's ERDAS IMAGINE software:

Imported individual JPEG2000 files to *.img

Combined tiles using Mosaic Pro and histogram matching color correction

Grids were recorded at different times, resulting in color imbalance. Color correction was needed for consistent image analysis.

JUNE 1ST, 2011

TORNADO

A 2015 STUDY IN

REMOTE SENSING

SPRINGFIELD

MONSON

STURBRIDGE

February 2015:

The author captures the snow-covered path of destruction from above Sturbridge, MA

Resulting in:

Several techniques can be used to better identify the path of damage:

False color RGB (near-infrared, red, green)

In the imagery's native resolution, individual trees which have been downed by the storm are clearly visible.

Subset to conventional shape (normal color RGB):

A standard Normalized Difference Vegetation Index (NDVI) produces a panchromatic image representing the ratio of red to near-infrared wavelengths reflected by vegetation. A darker area, indicating less vegetation, is still visible on the damage path after four years.

Image Analysis

A supervised classification with nine training signatures based on the above false color RGB produces a less clear result. Since the imagery was taken in early spring, it is difficult to differentiate between living deciduous stands and blowdown from the tornado.

However, isolating the layer identifying conifer stands reveals a distinct lack of this vegetation along the damage path.

...As does isolating the layer identifying bare earth or exposed pavement.

Conclusions

Of the three methods used above, the NDVI appears to produce the most distinctive results.

In addition to NDVI analysis, academic studies have also had good results using Primary Component Analysis (PCA) and Object-Oriented Approach to identify tornado tracks using before and after imagery.

Another application of remote sensing not addressed in this case study, satellite imagery and aerial photos are used by governments and the insurance industry to assess property damage after major storm events, such as a tornado.

A vector approximation of the tornado path

Credit: Tripline/Newsmappers, Google