samrae Posted October 4, 2012 Report Share Posted October 4, 2012 Hi there I am working with an NDVI image and was hoping to correlate the NDVI values with plant species richness, as there have been a number of studies that have used NDVI to predict species richness. However, in the various habitats I get different values: Forest* - medium values Pine plantation - medium values Riverine - high values Tea plantation - highest values *The forest areas are of old growth forest which should have the highest levels of biodiversity. But instead the riverine areas (and tea plantation) has higher values. I assume this is as NDVI is more closely correlated to new, live vegetaion (which emits more chorophyll). I was just wondering if I there would be any correlation to the species richness in this case then? And if there are other aspects, or insights that I may be missing... Thanks in advance! Quote Link to comment Share on other sites More sharing options...
yousef2233 Posted October 4, 2012 Report Share Posted October 4, 2012 1- i think you should look for Spatial cross-correlation 2- geographically weighted regression if you are not worry about lag 3- if you are not interested in spatial correlations, extract values in a specific grid location and use it in some statistics softs like minitab or spss 1 Quote Link to comment Share on other sites More sharing options...
3dbu Posted October 5, 2012 Report Share Posted October 5, 2012 is interesting to proof what is the process that followed for run NDVI index analyst and what type of soft used...the band combination and the general the process...regards Quote Link to comment Share on other sites More sharing options...
samrae Posted October 6, 2012 Author Report Share Posted October 6, 2012 Hi, thanks for the responses: @yousef2233: Thank you, for your suggestion but unfortunately at this time I dont have time to apply more methodological analysis. At this stage I would like to try to understand and explain the relationship of species richness - NDVI correlation (if it exists) at a conceptual level for my results. Or how what the NDVI values indicate about the different land cover classes? @3dbu: It was actually a very simple application using the NDVI function in ERDAS Imagine. There were actually some software problems with the band combinations, which the forum was really great about providing advice about (see http://www.gisarea.com/index.php?/topic/2577-ndvi-values-from-spot-imagery/) I should probably explain that this NDVI section fits as a small component to a larger thesis where I have looked at land cover change (hence the land cover classes above) and also indigenous knowledge in the area. Quote Link to comment Share on other sites More sharing options...
yousef2233 Posted October 6, 2012 Report Share Posted October 6, 2012 @samrae, unfortunately, i have no special experiences in remote sensing, but i have an idea, if you are looking for such correlations in a conceptual level, normalizing, then classifying those images and then using dot product maybe help Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.