Frequently Asked Questions (FAQ) / Common Issues

Map Creation

In general, we cannot offer assistance with creating overview maps or similar, as the methodology for map creation using GIS is taught in the curriculum for both Bachelor's and Master's programs and is required by accreditation standards. Furthermore, no external assistance is allowed for examination-related work. However, exceptions exist in Master's programs for those who did not learn GIS methodology during their Bachelor's studies. Please contact your supervisor directly in such cases and then reach out to us if necessary.

GIS

1. Ensure you have exported the table as a CSV file or XLSX file (the workflow below explains using a CSV file).

2. Check the coordinate system in which the coordinates in the table are provided (e.g., UTM, latitude & longitude).

3. Open the Data Source Manager (under the "Layer" menu).

4. Under the "Delimited Text" option, select the CSV file. Here, choose the appropriate columns for the X & Y coordinates.

   1. Ensure that the first row is always read as the "Header" or the names of the respective columns.

   2. Specify the corresponding coordinate system under "Geometry CRS."

5. Click "Add" at the bottom, and the points will be added to the project as a temporary layer.

Display Issues / Projection Problems (/CRS). The following problems occur:

You are likely using a projection based on latitude and longitude (Long/Lat). Examples include the standard WGS 84 projection (EPSG:4326) or ETRS89 (EPSG:4258). For local/regional analyses or tasks, always use a UTM-CRS, as these use the local geoid for the respective UTM zones, minimize distortions, and express lengths in meters. This allows unrestricted use of geospatial analysis tools in QGIS.

You likely used a global or area-preserving projection simultaneously for the project, the basemap, or the polygons. Transform all your layers into an angular projection or directly into the appropriate UTM zone and set this for the entire project as well.

Layers likely have different projections or reference systems (e.g., one in UTM, the other in LongLat). Ensure that all layers used share the same CRS as the project's CRS. You can use the Transform, Reproject, and Warp functions for this purpose; creating a new document can also resolve the issue if needed.

  • Warp is used for reprojecting raster data, transforming it from one CRS to another while correctly positioning the data.
  • Assign Projection changes the CRS of vector data but does not reproject the data. This means the coordinates are not altered but only translated into another CRS.
  • Translate keeps the CRS unchanged and merely shifts geometries in a layer by the specified units (e.g., by 10 m on the x-axis and 20 m on the y-axis).
  • Reproject layer creates a new layer with the specified CRS and reprojects all geometries into the new CRS, converting the coordinates of all objects.

Remote Sensing - Classifications (LULC)

This depends on your research question and/or hypothesis. For instance, if you want to determine land use in a study area and need to differentiate between various agricultural uses or are bound to predefined classes (examples: Hydrological Soil Groups/Curve Numbers or land cover classes based on state authority guidelines), a supervised classification is more suitable.

Short version: We recommend using the tools available in the C-Pool. Since the ArcGIS licenses expire in the winter semester 2024/2025, only ENVI will be available there. However, these tools are also available in QGIS (see: long version).  

Long version: For consistent results or when large datasets (e.g., a large watershed) are to be analyzed, the tools provided by ENVI (see: https://www.l3harrisgeospatial.com/docs/classification.html) or the Classification Wizard in ArcGIS (see: https://pro.arcgis.com/en/pro-app/latest/help/analysis/image-analyst/the-image-classification-wizard.htm) are recommended.  

Additionally, scripts and modules for the Google Earth Engine platform are now available for object-based classifications. More information can be found here:  

   - https://www.mdpi.com/2072-4292/12/22/3776  

   - https://github.com/SuzannaLin/GEE_LULC  

If you don't have access to these tools (e.g., no ArcGIS or C-Pool account is set up, or you need/want to work from home) or prefer open-source programs, you can use the GRASS GIS Plugin or the Semi-Automatic Classification Plugin (SCP) in QGIS.  

The R package "terra" is also recommended, allowing you to create both classified and unclassified LULC.  

   - [Documentation for unsupervised classification with Terra](https://rspatial.org/rs/4-unsupclassification.html)  

   - [Documentation for supervised classification with Terra](https://rspatial.org/rs/5-supclassification.html)
Unfortunately, we cannot provide technical support for SCP. This decision was made because the plugin is poorly integrated into QGIS. Due to infrequent updates and adjustments to new QGIS versions, bugs have become more frequent, such as the limited search mask no longer working or image metadata not being processed. In some cases, specific bands are no longer downloaded.  

Additionally, the plugin lacks comprehensive documentation, with only an FAQ available. Many error messages are not explained, making proper troubleshooting impossible. The usability also prevents processing large datasets, as entering spectral resolutions is time-consuming, super- and subclasses for supervised classification must be reassigned, and editing spectral signatures is not possible without re-importing files. There have also been instances where created training regions could no longer be loaded.  

Therefore, the use of SCP is not recommended. Instead, the programs and modules mentioned in the previous section (especially ENVI) should be used.

The most commonly used tool for this is eCognition. However, no free licenses are available for students for this commercial program. It is possible that you might gain access as part of a seminar, Bachelor's, or Master's thesis.  

Alternatively, the R library "segoptim" can be used free of charge. More information can be found here.

ENVI also offers suitable tools.

Wird geladen