Jgrocha: /* Calculating NDWI */

2018-06-11T21:26:29Z

Calculating NDWI

Jgrocha: Created page with "=== Calculating indexes === The Band calc allows for the raster calculation for bands (i.e. calculation of pixel values) using NumPy functions. Raster bands must be already l..."

2018-06-11T20:23:48Z

Created page with "=== Calculating indexes === The Band calc allows for the raster calculation for bands (i.e. calculation of pixel values) using NumPy functions. Raster bands must be already l..."

New page

=== Calculating indexes ===

The Band calc allows for the raster calculation for bands (i.e. calculation of pixel values) using NumPy functions.
Raster bands must be already loaded in QGIS. All rasters are listed in the Band list table.

Each raster in the has one variable and one name assigned. We can use the variable or the name. If we use the name, it must be written between double quotes.

When we have a band set defined, it is easier to identify each of the bands with "bandset#b1", "bandset#b2", etc. Besides that, we can also identify some special bands in the Band set by its name:
* "#BLUE#": the band with the center wavelength closest to 0.475 μm;
* "#GREEN#": the band with the center wavelength closest to 0.56 μm;
* "#RED#": the band with the center wavelength closest to 0.65 μm;
* "#NIR#": the band with the center wavelength closest to 0.85 μm;

Operators and functions

Besides normal operations, like + (plus), - (minus), etc, and some functions like sin, cos, etc, we can use and [https://docs.scipy.org/doc/numpy/reference/routines.math.html NumPy] function.

==== Calculating NDVI ====

Let's create the [https://en.wikipedia.org/wiki/Normalized_difference_vegetation_index NDVI] index to give examples of the band set expressions.

To create a new raster with NDVI values, we can use the default expression (available in the dropdown box):

<syntaxhighlight lang="text">
( "#NIR#" - "#RED#" ) / ( "#NIR#" + "#RED#" ) @NDVI_20180606_a
</syntaxhighlight>

The above expression uses the special names "#NIR#" and "#RED#" to select the bands from the Band set.

[[File:Captura de ecrã de 2018-06-11 21-44-27.png]]

The same index could be computed with the expression:

<syntaxhighlight lang="text">
( "bandset#b7" - "bandset#b3" ) / ( "bandset#b7" + "bandset#b3" ) @NDVI_20180606_b
</syntaxhighlight>

The above expression works for our Sentinel-2 Band set. Sentinel-2 NIR is band 08, so it is the 7th band in our band set.

NOTE: the last expression token <syntaxhighlight lang="text" inline>@NDVI_20180606_a</syntaxhighlight> defines the layer name. It is convenient to save the TIF output with the same name.

After running the calculation, a new raster layer is added to QGIS. Check the pixel values. They should be in the range <syntaxhighlight lang="text" inline>[-1, 1]</syntaxhighlight>.

[[File:Captura de ecrã de 2018-06-11 22-01-24.png]]

==== Calculating EVI ====

EVI index is anothe rpredefine index. Just select it from the drop down combobox.

==== Calculating NDWI ====

NDWI stands for Normalized Difference Water Index. There are authors using this index to '''Monitoring Crop Field Soil Moisture'''.

Let's create the NDWI for several Sentinel-2 images to have some insights about soil moisture.

@@ Line 50: / Line 50: @@
 ==== Calculating NDWI ====
-NDWI stands for Normalized Difference Water Index. There are authors using this index to '''Monitoring Crop Field Soil Moisture'''.
+NDWI stands for '''Normalized Difference Water Index'''.
 Let's create the NDWI for several Sentinel-2 images to have some insights about soil moisture.

Raster algebra: calculating indexes - Revision history

Jgrocha: /* Calculating NDWI */

Jgrocha: Created page with "=== Calculating indexes === The Band calc allows for the raster calculation for bands (i.e. calculation of pixel values) using NumPy functions. Raster bands must be already l..."