Friday, July 4, 2014

Spatial Analysis - Sifting Statistics


This is Part III in a series of articles on Wildfires in the desert region of Gold Butte in North Eastern Clark County, NV. To read Part II click the following link: Defining the Study Area:


In the field of Spatial Analysis and Statistics you use geographic data, which is data that has a fixed location within the real world, to find trends and correlations between the data. The goal of my fire analysis project is to find try and find correlations between the fire area and the rest of the gold butte region to create a risk index to identify the most at risk areas in Gold Butte that haven’t yet burned but are likely to burn. Once this is complete I will then document those areas current habitat for future restoration plans. I will also look for ways to protect these desert ecosystems from the devastating consequences of a wildfire event.



In step 1 I defined the sampling area as the boundary of the Forks and Tramp fire within the Gold Butte Region and presented the summary five different datasets including soil, geology, vegetation types and landforms or slope classifications within the burned area (sample).

 In step 2 I defined the study area to which I would scale my analysis to. I then presented the summary results of the same datasets that I presented for the sample area in step 1.

In Step 3 I am looking for recurring patterns and relationships between the burned area and the total study area. In this step I am looking for scale invariance or the lack of scale invariance to try and see if I can determine why the fire burned where it did through analyzing the results of both step 1 and step 2.



For example, if all of the data from the sample area scaled perfectly to the study area, all that it would tell me is that this fire burned consistent with the statistical distribution across the total study area.  If 30 percent of the study area was made up of creosote and 30 percent of the burn area was also creosote the only inference that could be made is that the fire burned consistent with the overall makeup of the study area. However if 30 percent of the study area was made up of creosote but 5 percent of the burn area consisted of creosote then we could deduce that the creosote vegetation is not as susceptible to fire.  The following is the breakdown of the previously defined data:

Landform (Slope Classifications)
Looking at the total Landform distribution across the entire study area you can see that the gently sloping ridges and hills make up the majority of the study area. However that same classification only makes up 20% of the burn area meaning that this landform classification is not as susceptible to fire as other types.  The landform types that have the smallest percent of the total study area but the largest percentage of the burn area are the indicators of the most susceptible landform types:
·         very dry steep slopes
·         very moist steep slopes
·         hot aspect scarps, cliffs, canyons
·         cool aspect scarps, cliffs, canyons





Geologic Types
Again here I am looking for the geologic types that have the smallest percent of the total study area but the largest percentage of the burn area.  The most susceptible geologic types:
·         very dry steep slopes
·         very moist steep slopes
·         hot aspect scarps, cliffs, canyons
·         cool aspect scarps, cliffs, canyons





Soil Types
After analyzing the soil types the indicators are not as significant as other data is at showing correlation however there are still relationships that exist that will help add to the modeling. Again I am looking for the types that have the smallest percent of the total study area but the largest percentage of the burn area.  The most susceptible soil types are:
·         Water (this is the land that has been exposed by the drop in lake water levels)
·         lithic torriorthents-rock outcrop-lithic and deep calciorthids
·         deep and shallow paleorthids-calciorthids-haplargids






Vegetation Types
Again I am looking for the types that have the smallest percent of the total study area but the largest percentage of the burn area.  The most susceptible Vegetation types are:
·         Artemisia tridentata ssp. (tridentata, wyomingensis) – This type was only found in burn area
·         Inter-Mountain Basins Montane Sagebrush Steppe
·         Great Basin Pinyon-Juniper Woodland
·         Inter-Mountain Basins Big Sagebrush Shrubland
·         Agriculture-Cultivated Crops and Irrigated Agriculture
·         Mojave Mid-Elevation Mixed Desert Scrub







Vegetation 2
Again I am looking for the types that have the smallest percent of the total study area but the largest percentage of the burn area.  The most susceptible Vegetation types are:
·         Juniper II
·         Mountain shrub

·         Blackbrush