Analyses

To: Class // From: George // Date: 2010 Nov 5

In response to what we learned at this morning's session (attended by Kevin, Nyeema, George, Melissa) and to requests for clarification by Nyeema, I've revised slightly the research questions that we are working on. They appear on the top of our scenario ideas page - I pushed all of our earlier discussion down.

Notice that I split them into two groups - questions about the correspondence of ES and biodiversity; and the question about payment for ecosystem services v tax value.

Here's our status ...
 * We have data on areas of conservation concern and on the location of individual species of ecosystems of conservation concern
 * We have developed procedures for answering all the pieces of (the new) Q1.
 * We are still working on getting additional biodiversity data from the NC GAP program - we can do some analysis with these data, but for others we're still trying to develop a procedure that works.
 * Kevin & Nyeema are working on this.
 * All data are available, 2/3 of procedures ready, but the last 1/3 or procedure is going to be difficult


 * We have biophysical data for pollination that are ready for stacking
 * There will be no $$ value for pollination - the model does not deliver that and we will not deal with it
 * We think pollination is done.


 * The carbon biophysical and $$ were not available to us this morning.
 * We've been assured that they exist and will be posted shortly by Meg to the layers files page
 * They are both ready for stacking.
 * We think carbon is done.


 * We noticed what appear to be anomalies in the biophysical nutrient retention layers
 * The forested areas showed the lowest levels of nutrient retention, despite having the deepest soils and forest cover - we did not expect and do not understand the patterns of nutrient retention we saw.
 * There were several areas of nodata - most were water, but some were in random urban areas
 * Nutrient team is still working on cost values - a few parameters are still unclear
 * Cost layers not yet available
 * Nutrient team is evaluating these issues and will report Monday

To: Kevin, From: george On: 2010 Nov 2 Re: Prelim analyses

Once the layers files are up can you please do some analyses to move us closer to our sketch graphs ( [|GraphSketches.pdf] )?

Question 1 - as there is no time series here for now, we just need to show the state of the landscape with respect to ecosystem services. HEre's what I think makes sense, but feel free to differ.
 * For each biophysical model, I think we want a map with a mono greyscale coloramp with 0 white and max value black.
 * Add all 3 normalized biophysical models to get a total biophysical ES - should range 0-3, and do a mono greyscale of that. (This is the summed normalized layer.)

Question 2 - How much biodiversity is included by ES areas?
 * Quartile the summed normalized layer, but keep 0 as a separate entry.
 * Take the highest quartile and figure out what portion of the significant natural heritage areas is included within it. This is going to be a raster / vector operation of some kind (SNHA is vector).
 * Then do it again for the top two quartiles combined, then the top 3.
 * Now for the element occurrences, we want to know about element occurrences within 500FEET (152m) of the top quartile ES areas (same places used for the SNHA). The 500FEET comes from earlier work Hess etal 2008 that we'd like to be consistent with. We need a list of all the EOs that meet this criterion. The list has to include the name of the species or ecosystem. From that list, we can generate completeness and representativeness stats. Again, repeat for top two quartiles combined, then the top 3.

Question 3 - How much ES is included in biodiversity areas?
 * This is kind of the flip-side of Q2.
 * Use the quartiled summed normalized layer, clip to the SNHA layer and dump the attribute table. This should tell us how many pixels of each quartile (plus 0) are contained in SNHAs. From that, we can create the graph sketched for Q3

Question 4 - Tax stuff
 * I think the main thing to do here is figure out how to deal with the tax data. I'm not sure we want to look at ES vs tax within the built areas, but I'm not sure how we'd mask them out (or if we should). We need to find a way to deal with the sliver problem you noted. Once we have a good layer, the division should not be hard.