pappmarci95
commented
4 years ago Hi, Their method is based on a regular grid placed randomly on the layer (so every two points can fall in different cells). The grid must contain 4^m cells (2^m in one line), where m = log4(sample number). The cells are then ordered quadrant recursively. We create a line with segments for each cells placed by the previous order. This way the two-dimensional space is mapped to one-dimensional and we can sample that systematically. Inside every line segment there are subsegments representing the polygon parts in the grid cell. Their size is proportional to the ratio of the polygon part and the cell size and they are randomly ordered. This method creates evenly spread sample locations in the layer, but still have some randomness. Although it has good spatial properties, its greatest advantage is how easily unequal probability sampling can be achieved by changing the length of the line segments, but unfortunately this feature is not yet implemented in the plugin. Besides these features the method is easily applicable to the QGIS environment, and doesn’t need complex programming solutions. You can read more in the article you requested, in the site: https://cfpub.epa.gov/ncer_abstracts/index.cfm/fuseaction/display.files/fileID/13339 I also recommend the book Design and Analysis of Long-term Ecological Monitoring Studies, where you can find a more detailed description of the method between the pages 126 and 150 (the R application of the method also cites this book), if you are interested. You can find the details of the book for example here: https://www.cambridge.org/core/books/design-and-analysis-of-longterm-ecological-monitoring-studies/508A10FEE39E7E93EF07B005D06952F5
tobwen
Could you please write a short introduction about the paper from Stevens and Olsen (2004)? What makes it so special? Do you have some illustrations? Could you also please share the paper?