Irish landcover change 2009-2012- New LUCAS data published use with some caution

The preliminary results for the EEA/Eurostat LUCAS frame area survey for 2012 for Ireland have been published. You can find all the details and data here but in short the landcover at point on a 2km grid covering the country were recorded on a field trip. 4000 or so points and they can be directly compared to a similar survey in 2009. I’ll be working on this data for a while but I thought I’d show you a danger warning first.

The graph below shows the comparison between 2009 and 2012 at an aggregate scale for Ireland and you can see that there is little change of interest (the numbers are the numbers of points recorded)


Nothing remarkable there but when you look at point by point change there is a huge amount of churn. The table below looks at points in 2012 with the same ID as 2009 and compares Landcover code.


obviously 25% of built land in 2012 was not grassland in 2009, so something needs to be explained. Clearly there is a high degree of spatial uncertainty in these repeat observations (not all by any mean are visited direct many are viewed from afar or via remote sensing).

The main reason for this is that the locations of the target points are not always visited. The first point in the list is shown below with the target in Red, the location from where the observation was taken in 2012 in Blue and the location in 2009 in yellow(the imagery is Bing Maps from 2011):


Given the quoted uncertainty 2009 was obviously observed from the road. the observation in 2012 is H12 Peat Bogs and in 2009 D20 , Shrubland without Trees. Indeed if we compare the locations of 2012 observation with 2009 (remember supposedly observing the same spot), we get the histogram below:



So only half of the points in 2012 that code to the same location as 2009 were observed from within 50m of the original observation point (neither of which need be the actual target). So I’d be wary, for now, of looking at landcover flows – overall change statistics at National level is the designed use – better stick to it unless some guidance from Eurostat is published.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s