[GIS] How to do Network Kernel Density Estimation Analysis in ArcGIS for Desktop

arcgis-10.0arcgis-desktopkernel densitymodelbuildernetwork-analyst

I am looking at doing some analysis of road traffic collisions. Obviously collisions occur on stretches of roads and I was researching the best methodology to take into account the network element of it. The usual hot-spotting (Kernel Density) smooths things out too much, and I found a few articles on Network Kernel Density Estimation which is supposed to be more representative as it calculates the distance between events via the network rather than using euclidean distances. We are using ArcGIS Basic/Standard 10 at the moment. We do not have Spatial Analyst licence but we have Crime Analyst and Network Analyst.

Anybody has done this before in ArcGIS? I've googled it I can't find any ready made tools – so I was wondering if anybody had a script I could use? Or any idea on how I could do it in a model?

Best Answer

There is a set of add-on ArcGIS Desktop tools (versions 9- 10.2) using network kernel density techniques developed by Professor Okabe available here. http://sanet.csis.u-tokyo.ac.jp/.

However they are only licensed for academic use. I used them on my MSc thesis and they worked really well. I guess you could enquire about commercial implementations.

According to the documentation there is an open source implementation in GRASS (v.kernel).

Related Solutions

[GIS] Heatmap using Kernel density estimation

I think you need to replace your radius parameter with a number.

Incase this might help for future reference, what I usually do when I need to know what parameters I have to set, is to check the "help" details of the algorithm within Python Console (Plugins > Python Console). Type the first 2 lines below:

>>>import processing
>>>processing.alghelp("saga:kerneldensityestimation")

ALGORITHM: Kernel density estimation
        POINTS <ParameterVector>
        POPULATION <ParameterTableField from POINTS>
        RADIUS <ParameterNumber>
        KERNEL <ParameterSelection>
        TARGET <ParameterSelection>
        OUTPUT_EXTENT <ParameterExtent>
        USER_SIZE <ParameterNumber>
        USER_GRID <OutputRaster>

KERNEL(Kernel)
        0 - [0] quartic kernel
        1 - [1] gaussian kernel
TARGET(Target Grid)
        0 - [0] user defined

[GIS] R Spatial Kernel Density Estimation

I have also been looking for a proper way to perform a weighted bivariate kernel interpolation. The code below worked for me:

# Download an example dataset - those are tree logs in a 100x100m plot. I used the volume of log, as weight.

test <- read.csv("https://dl.dropboxusercontent.com/u/39606472/R_rep/test.csv")
require(ks)

# Evaluate effect of tree felt out and in the plot. By effect I mean a mixture between trunk volume and distance to tree on a regular grid. xmin/xmax = plot area, eval.points= where I want the effect to be evaluated

    kernel <-kde(x=test[,c("x1","y1")], xmin=c(-20,-20),xmax=c(120,120),eval.points = expand.grid(x=seq(5,95,10), y=seq(5,95,10)),w=res$vol)
        IDW <- data.frame(x=kernel[[2]]$x,y=kernel[[2]]$y,z=kernel[[3]])
        plot(test$x1,test$y1,cex=log(test$vol),xlim=c(-20,120),ylim=c(-20,120))
        image(IDW,add=T)
        points(test$x1,test$y1,cex=log(test$vol))

Best Answer

Related Solutions

[GIS] Heatmap using Kernel density estimation

[GIS] R Spatial Kernel Density Estimation

Related Question