California Drought

Wet Year (2011) vs. Dry Year (2014)

ES 771 Remote Sensing Fall 2014
Kyle Jackson & Calif Tervo

Image Acquisition

Google Earth was used to find the latitude and longitude of location of interest.

We searched the United States Geological Survey (USGS) Global Visualization Viewer (Glovis). (Collection> Landsat Archive>Landsat 4-present), and input the latitude and longitude of a location of interest.

We looked for two or three locations with good quality images for September 2011 and September 2014 and with the location contained on a single row and path.

Files of potential locations were downloaded to “downloads”, moved to working file and extracted using 7-zip; then imported into IDRISI Selva. (File > Import > Government > GeoTiff).

After looking at the images, the locations of Silver City, Spanish Flat, and Santa Cruz Mts. were chosen.

The following images were obtained:

Silver City (southern Sierra foothills)
WRS-2
Row: 42
Path:35
Lat: 36.0
Long: -119.5

ID: LT50420352011263PAC01
CC: 0%
Date: 2011/9/20
Qlty:9
Product TM L1T

Subscene:
Approximate sub-scene location: latitude:36.0, longitude -119.5
Upper Left Column: 6162
Upper Left Row: 2013
Lower Right Column: 7160
Lower Right Row: 2929

LC80420352014255LGN00
CC: 0%
Date: 2014.09/02
Qlty:9
Product: OLI_TIRS_L1T

Subscene window to:
Upper Left Column: 5782
Upper Left Row: 2280
Lower Right Column: 6997
Lower Right Row: 3291

Spanish Flat (middle Sierra Foothills)
USGS Glovis
Collection> Landsat Archive>Landsat 4-present
WRS-2
Path: 43
Row: 33
Lat: 38.9
Long: -120.2

ID: LE70430332011262EDC00
CC: 0% Date: 2011/9/19
Qlty: 9 Product: ETM+ L1T

ID: LE70430332014254EDC00
CC: 0% Date: 2014/9/11
Qlty: 9 Product: ETM+ L1T

Approximate sub-scent location: latitude: 38.35, longitude. -120.35
Window: Col.: 3200, Row: 5400; Col.: 3850, Row: 6050

Santa Cruz Mts.
USGS Glovis
Collection> Landsat Archive>Landsat 4-present
WRS-2
Row: 44
Path 35
Lat.: 37.5
Long: -122.1

ID: LE70440342011269EDC00
CC: 9% Date: 2011/9/26
Qlty: 9 Product: ETM+ L1T

ID: LE70440342014277EDC00
CC: 0% Date: 2014/10/4
Qlty: 9 Product: ETM+ L1T

Approximate sub-scene location: latitude:37.20, longitude -122.10
Window: Col.: 3500, Row: 4200; Col.: 4165, Row: 4800

back to top

Methodology

For each location, all images retrieved from Landsat were cropped to the study location (Reformat > WINDOW).

Note: Band numbers listed below are equivalent of Landsat 4, 5, and 7. For Landsat 8, bands 1-5, 1 was added to the Landsat 8 band number.

For each location, some or all of the following images were processed:

1) Standard “Natural Color” composite images were created (Display > COMPOSITE).
The following band combination was used for Landsat 7 images (for Landsat 8, bands 1-5, 1 was added to get the COMPOSITE band):
Blue: Band 1 (blue)
Green: Band 2 (green)
Red: Band 3 (red)
Contrast stretch type > Linear with saturation points
Output type > Create 24-bit composite with original values and stretched saturation points.
Percent to be saturated> 1.0.

2) Standard “False Color” composite images were created (Display > COMPOSITE).
The following band combination was used for Landsat 7 images (for Landsat 8, bands 1-5, 1 was added to get the COMPOSITE band):
Blue: Band 2 (green)
Green: Band 3 (red)
Red: Band 4 (NIR)
Contrast stretch type > Linear with saturation points
Output type > Create 24-bit composite with original values and stretched saturation points.
Percent to be saturated> 1.0.

3) NDVI Vegetation Index image was also created for all four study areas.
Before NDVI images were created, bands (2-5) were corrected for haze. To correct for haze the minimum value was set to “1” by adding or subtracting a value using the SCALAR function (GIS Analysis > Mathematical Operators > SCALAR).

Once the bands were corrected, NDVI images were created using VEGINDEX (Image Processing >Transformation > VEGINDEX > NDVI) was selected and the following band combination was used:
Red: Band 3 (haze corrected)
Infrared: Band 4 (haze corrected)

Normalized Difference Vegetation Index (NDVI) computed by:
NDVI = (NIR - Red)/ (NIR + Red). This produces a value between 1 and -1.

4) NDVI False Color Image using NADI was created

First: the NDVI image values of 1 to -1 were normalized to 0-255 values by the following: The SCALAR function was used to reformat the NDVI Vegetation Index image by adding “1” ((GIS Analysis > Mathematical Operators > SCALAR). NDVI-34plus1 and multiplying by “125” NDVI-34plus1times125. Next, the CONVERT function (Reformat > CONVERT > Output data type: Integer, Output file type: binary, with rounding >OK, was selected. The NDVI False Color composite image was used using the NDVI image. (Display > COMPOSITE.

The following band combination was used:
Blue: Band 2 Haze Corrected (green)
Green: NDVI-34Converted (NDVI)
Red: Band 5 Haze Corrected (Mid-IR)
Contrast stretch type > Linear with saturation points
Output type > Create 24-bit composite wit original values and stretched saturation points
Percent to be saturated> 1.0.

5) Vegetation Ratio was created:
Once the bands were corrected, NDVI images were created using VEGINDEX (Image Processing >Transformation > VEGINDEX > Ratio) and the following band combination was used:
Red: Band 3 (visible red)
Infrared: Band 4 (near infrared)

6) Mid-infrared False Color Composites:
Once the bands were haze corrected, (Display > COMPOSITE).
The following band combination was used:
Blue: Band 2 (green)
Green: Band 5 (near-infrared)
Red: Band 7 (infrared)
Contrast stretch type > Linear with saturation points
Output type > Create 24-bit composite with original values and stretched saturation points.
Percent to be saturated> 1.0.

Outputed the processed files as .jpg using JPGIDRIS.

back to top