Lab 03 - Land Use/Land Cover Classification Mapping

Lab 03 was very enjoyable. Lecture and lab were on the topic of land use and land cover classification from aerial photo interpretation. Land use considers the human uses on the land, i.e. industrial complexes, airports, or houses, while land cover is a biophysical description of the Earth's surface, i.e vegetation and water bodies. I would guess almost everyone who has done GIS analysis or mapping has used an LULC dataset at some point. I know I have this year in more than one of my classes so it was fun to create my own LULC map.

Observing the surface of the earth through aerial and satellite imagery, particularly over time,  can be used for natural resource inventories, urban planning, LULC change analysis, military operations, and more. The LULC process starts with a research question or need, a classification scheme is chosen based on that, photos and other resources are gathered, interpretation begins, after which the classification polygons are verified for accuracy, and reports are written upon approval.

In this lab, we were provided an aerial photo of Pascagoula, MS and assigned the Level II classification scheme. I first spent a little time measuring areas I thought might fall into a classification code to help me determine a minimum mapping unit (MMU). This would help define how far to break down the features into classification polygons. From there I began to work my way from large areas, like the bay and residential mainland into more detailed interpretation. I found it easiest to create large polygons first and then use the Editor Clip tool to clip out smaller polygons. I used the interpretation elements we learned in Lab 02 to guide my classifications. For example, looking at shape, color, pattern, and shadows to find commercial storefronts, different types of trees, or high fences surrounding an area that could be industrial. As I was only working down to Level II classification there are some areas on the map, like schools, could be drawn into their own code class if going up to higher levels of classification. There were also a few areas I chose to list as mixed codes as I was staying around a MMU of 6 ha, such as not defining all the streams and rivers within a wetlands area. Again, for a more detailed map those areas could be broken down into smaller polygons and more distinct classes.

After creating my polygons and labeling them, I added transparent color to more easily differentiate between the codes and still see the underlying features. Similar to our last lab, there is a good deal of personal judgement in creating an LULC, however the classification levels, pre-existing maps, collaboration, experience, and past standards can help reduce human variance in interpretation.

Land use, land cover map created through aerial photo interpretation using 6 ha as the MMU and a Level II classification scheme.