Vegetation - Central Valley Flood Protection Planning Area Update - 2016 [ds2890]

The results of this analysis can be found in Table 4. Riparian acres totaled 281,033 in 2009 and 303,823 in 2016, showing an increase of 22,790 acres (8%). Non-riparian acres were 291,582 in 2009 and 264,568 in 2016, showing a decrease of 27,013 acres (9%).

Agriculture covered 1,753,706 acres in 2009 and 1,757,906 acres in 2016, showing an increase of 4,200 acres (<1%). Urban areas covered 241,002 acres in 2009 and 244,155 acres in 2016, showing an increase of 3,153 acres (1%). Change detection is a modern concept where two maps covering the same area but based on different imagery years are compared using ArcGIS tools.

The output tells us what has changed and, with further inquiry, one can determine specific changes from one mapping unit to another. For example, a wetland area mapped as Managed Marsh in 2009 had small willow trees that were not tall enough or dense enough to be considered a tree stand. By 2016, these willow trees had matured enough over 7 years to be classified as a willow stand.

This example occurred in the Butte Sink area west of the Sutter Buttes. Changes in a remap effort can occur due to several factors, some due to physical change and some due to non-physical change. Examples of factors that represent physical change are: stand maturation or decadence, fire, erosion, deposition, anthropogenic disturbances such as scraping/clearing, urban development, agricultural development, and restoration.

Non-physical changes include: improvement in imagery quality, incorrect original mapping, imagery shifts, improvement of older linework, and change in level mapped (classification). Imagery quality continues to improve with time, which allows the photointerpreter to better determine which species are present and in what amounts to determine its best classification. A photointerpreter could also disagree with the previous mapper’s decisions and may change the classification in the remapping effort.

A more subtle map change, which does not affect the acreages significantly, is when the photointerpreter improves old linework to make it better fit the classification type visually. Also, when the NAIP imagery is ortho-rectified, in some years it does not perfectly align with the year it is being compared to and linework needs to be shifted/redrawn. Lastly, a photointerpreter may have remapped an Alliance at the Group level due to lack of confidence at the Alliance level, and vice versa the interpreter may have mapped an Alliance that was previously mapped at Group level due to better confidence at that level.

All of these factors should be taken into consideration when using and comparing these data to previous efforts. Please see the report here nrm.dfg.ca.gov/FileHandler.ashx?DocumentID=184992 for more information on these data and the change detection results.