Master Course

Photogrammetry and Geoinformatics

Prof. Dr. M. Hahn

Prof. Dr. E. Gülch

Tutorial

Appendix to the Assignment

Interpretation of Sentinel and Landsat Imagery

using ERDAS Imagine

App. 1 Tutorial on spectral signatures related to task 2

Investigations using the Spectral Profile of a pixel over all bands

(Multispectral, Utilities, Spectral Profile---create a new profile point +)

Observe the differences of the pixels values in the bands for different land cover types.

The chart can be changed/improved (Edit Legend, Edit Options) and points can be deleted
(Edit, Delete Plots, Check on the Colomn “Delete Plot”, then OK).
Points can be moved to other positions.

Tutorial_Sentinel_Landsat_updateApril2022.docx update March, 2022Baur 1

Master Course Photogrammetry and Geoinformatics Prof. Dr. M. Hahn

App. 2 Tutorial on histograms related to task 2

Investigations using the histogram of different land cover classes

This can be done in ERDAS Imagine using the AOI_Layer and Subset (Drawing—Polyon)
but it is suggested to use the Signature Editor which you know from the Classification
Exercise.
(Raster---Supervised –Signature Editor)

Here you can visualize several histograms of areas in one plot window.

App. 3 Tutorial on NDVI and visualisation of results related to task 3

Raster – Unsupervised - NDVI

To visualize the NDVI open the NDVI-File as Pseudo Color File (Open Raster Layer, choose
Pseudo Color) or as Image Chain (Open –Raster as Image Chain).

Change colors if the file is a pseudo raster file the select some rows and change the color
(Table – Show Attributes – Criteria – selection criteria for rows opens)

Change colors as Image Chain (Panchromatic – Image Chain – Pseudocolor – Tab

Pseudocolor - choose a Color Table – Open the Stretch Panel for Enhancements)

Color Table NDVI Natural Color is used in this example.

Tutorial_Sentinel_Landsat_updateApril2022.docx 09.06.23-MH 2
Master Course Photogrammetry and Geoinformatics Prof. Dr. M. Hahn

App. 4 Tutorial on burnt area analysis related to task 4

1. Display the dataset Narbonne, Sentinel2a (post-fire) to find out the burnt area. A
shortwave infrared band combination of bands (12, 11, 8a) is used for mapping of fire scars
or burnt areas. (Attention to the band numbering in ErdasImagine)

An example of a burnt area in

which there is still a fire to be
seen shows the following
figure:

Since we now know where the fire took place, we want to investigate which kind of
vegetation was at this place, whether this vegetation was under stress, and how high the
temperatures had been before the fire. For that purpose we select an area of interest with the
burnt area in the center and agricultural fields, forest, the urban area of Narbonne, lakes and
the ocean in the neighborhood.

2. Calculate and visualise the NDVI from Landsat 8. (Raster – Unsupervised – NDVI).
- see tutorial above

Open the result as Pseudo Colour Image that you can set colours (not as Grey Scale)
(Raster – Table – Show Attributes)
Hints: Home – Inquire (show the NDVI in a certain pixel), use more views with the Landsat8
and the Sentinel2a data set and link the views and equalize the scales

Check the NDVI values, find for the given land cover classes proper ranges for the NDVI
values and fill them into the table below.

Define your own colour table (Show Attributes – Select Criteria)

Note: There is an “undo” function.

Tutorial_Sentinel_Landsat_updateApril2022.docx 09.06.23-MH 3
Master Course Photogrammetry and Geoinformatics Prof. Dr. M. Hahn

Table: Allocation of NDVI values to land cover classes and defining colours

NDVI Color Select Criteria

Range of all -0,36 – 0,68 Note: these NDVI
values-whole values are examples
image file
Range for most of 0,14 – 0,40
the values- whole
image file
Ocean, Lake -0,03 – 0,14 white $"Value" < 0.15

Urban Area, 0,2 – …. white

Roads
Agricultural Area 0,27 – 0,28 Light $"Value" > 0.21 and
green $"Value" < 0.28
Low vegetation, 0,16 – …. orange $"Value" > 0.15 and
fields $"Value" < 0.21
Forest 0,31 – 0,36 Dark $"Value" > 0.28
green

3. Calculate the moisture stress index MSI from Landsat 8, use the MSITIR-File (9 bands)
as input file (Raster – Unsupervised – Indices)

Example MSI

Open the result as Pseudo Colour Image (not as Grey Scale) and evaluate the MSI values
(Raster – Table – Show Attributes) in different areas in the AOI.
(Home - Inquire cursor)
For a fast and easy visualization of the MSI result a prepared colour table (Color Table BYR)
of ERDASImagine can be used (Open –Raster as Image Chain…..)

Tutorial_Sentinel_Landsat_updateApril2022.docx 09.06.23-MH 4
Master Course Photogrammetry and Geoinformatics Prof. Dr. M. Hahn

Fill out the following table in which you write the range of the MSI values for the given land
cover classes.
MSI Remarks
Range of all 0,24 – 3,41 Note: These MSI values are
values-whole examples
image file
Range for most of 0,52 – 1,11
the values- whole
image file
Ocean, Lake 0,89 – 0,92

River 0,51 – 0,65

Agricultural Area 0,90 – 1,10

Urban Area, 0,90 – 1,12
Roads
Forest 0,61 – 0,73

Burnt Area 0,75 – 0,87 – 1,1

4. Calculate the Brightness Temperature from Landsat 8

For this calculation use the given Spatial Model L8_Single_Them2Tem-v15-1-8.qmdx, the
metadata from file LC81970302016179LGN00_MTL_originalmetafile.txt and the Landsat
band 10 (TIRS 1).

Toolbox -Spatial Model----Launch Spatial Model (select the model and set the parameters)

Evaluate your results (Raster – Pseudo Color –Table –Show Attribute…..) and colourize the
hot areas, e.g. orange from 26 – 29 degrees, red above 29 degrees.

Tutorial_Sentinel_Landsat_updateApril2022.docx 09.06.23-MH 5
Master Course Photogrammetry and Geoinformatics Prof. Dr. M. Hahn

App. 5 Table: Modified Anderson Land Use/Land Cover Classification

Modified Anderson Land Use/Land Cover Classification used for the USGS National Land
Cover Dataset. (https://www.e-education.psu.edu/natureofgeoinfo/c8_p19.html)

Level I Classes Level II Classes

11 Open Water
Water
12 Perennial Ice/Snow
Developed 21 Low Intensity Residential
22 High Intensity Residential
23 Commercial/Industrial/Transportation
Barren 31 Bare Rock/Sand/Clay
32 Quarries/Strip Mines/Gravel Pits
33 Transitional
Forrested Upland 41 Deciduous Forest
42 Evergreen Forest
43 Mixed Forest
Shrubland 51 Shrubland
Non-Natural Woody 61 Orchards/Vineyards/Other
Herbaceous Upland Natural/Semi-
71 Grasslands/Herbaceous
natural Vegetation
Herbaceous Planted/Cultivated 81 Pasture/Hay
82 Row Crops
83 Small Grains
84 Fallow
85 Urban/Recreational Grasses
Wetlands 91 Woody Wetlands
92 Emergent Herbaceous Wetlands

Tutorial_Sentinel_Landsat_updateApril2022.docx 09.06.23-MH 6
Master Course Photogrammetry and Geoinformatics Prof. Dr. M. Hahn

App. 6 False colour band combinations of Landsat 8

https://blogs.esri.com/esri/arcgis/2013/07/24/band-combinations-for-landsat-8/

“Landsat 8: While all of the bands from previous Landsat missions are still incorporated,
there are a couple of new ones, such as the coastal blue band water penetration/aerosol
detection and the cirrus cloud band for cloud masking and other applications. Here’s a
rundown of some common band combinations applied to Landsat 8, displayed as a red,
green, blue (RGB)”:

Natural Color 432

False Color (urban) 764
Color Infrared (vegetation) 543
Agriculture 652
Atmospheric Penetration 765
Healthy Vegetation 562
Land/Water 564
Natural with Atmospheric Removal 753
Shortwave Infrared 754
Vegetation Analysis 654

Tutorial_Sentinel_Landsat_updateApril2022.docx 09.06.23-MH 7