Some Very Cool Data
This post reviews LiDAR data available for the southern continent and the Dry Valleys in particular. From the raw data we derive a high resolution DEM.
I hate to say this, but some data are cooler than others (Ed. What a geeky thing to say…) and LiDAR data certainly fit into that category. We’ve talked about LiDAR earlier – safe to say this it is revolutionising geospatial data one DEM at a time. I’ve been doing a bit of reading on that rather large continent to our south and wondered if the LiDAR revolution had arrived on its shores. The short answer is yes.
With a bit of help, I found the Polar Geospatial Center (sic), which “provides geospatial support, mapping, and GIS/remote sensing solutions to researchers and logistics groups in the polar science community”. Be still my beating heart…! Under “Data + Services > Elevation” there’s a section on the McMurdo Dry Valleys LiDAR (could someone call a doctor please?) which includes this image:
WE NEED 50 CCs OF EPINEPHRINE STAT!
So, once I was revived, I followed this inviting looking link:
Some potentially nice looking stuff in here – tifs, maps, docs.
When working with LiDAR data a common file format is “las” files.
So that’s my next stop:
Lately I’ve been a bit interested it the Dry Valleys and the Wright Valley in particular and the image above suggested I look in a folder called “Northern_DV_South”. This folder contained no less than 359 las files and (gulp) 39.3 Gb of data. Okay, I couldn’t resist – don’t tell ITS – but I downloaded the whole lot using an FTP client:
Three days later (no, more like three hours really) I had these on my hot little hard drive (Ed. that sounds quite risque – can you tone it down please?). Each of these las files is a subset of the original data, broken down into rectangular tiles roughly 1 km by 1 km in size. This is quite standard for packaging up LiDAR data – this has its pluses and minuses but as a next step I needed to know where all these tiles are. I found one in the docs directly and downloaded it as a shapefile (there are 4062 tiles, shown below in ArcGIS Pro):
I’ll focus in on the upper end of the Wright Valley for these next stages – I’m aiming at creating a high resolution DEM.
In this image I’m zoomed in on the upper Wright Valley with one of the tiles selected:
I’m going to need to know which tiles relate to this area which will be a bit of a trick, given the complex file names. Also note the Pt_Count (number of LiDAR points) in the las file for this tile: 3,202,421! And that’s just one tile! Also note the Pt_Spacing (average distance between LiDAR points) of 0.5588 m. That’s a lotta points… So, to make a long story short, I’ve identified all the tiles I want to use for my DEM (225) and created a LAS Dataset – this is a data structure that makes it a lot easier to manage all those points – there will be 901,999,433 in total to work with – I’m worried I’m going to break something…
First step is to create my LAS Dataset using the aptly named “Create LAS Dataset” tool (who comes up with these names?). When displayed, it doesn’t look like much:
That’s because the points only display once we zoom in beyond a certain scale, around 1:5,000. The image below is around 1:2,000 to give you a sense of the density of points:
(Purple Rain anyone?)
Here it is in ArcScene:
This is nice for being able to visualise the data but to make it much more useful we really need is a raster DEM. I won’t go through the excruciating details (and several hours of processing time) but the raster output is shown below with a bit of transparency and a hillshade in a 3D Scene in Pro:
Sorry, but I just fell off my chair…there’s some exquisite detail in this DEM. Here’s Don Juan Pond, perhaps the most saline water body on the face of the earth:
And here’s a view looking east from the top of the valley roughly 20 km to the end of the dataset – awesome:
If you’re interested in playing around with this dataset, let me know and I’ll make it available on J: in all its glory.
As a final Antarctic note, a new 8m resolution DSM of all of Antarctica has recently been released and it’s stunning. This New York Times article gives you an overview.
Most data are cool, but some data are cooler than others – and these are particularly good examples.
C