Lately I’ve been getting indications of something very good. People are actually reading the metadata. This is great! I don’t know how many times I’ve heard someone dragging their feet when creating metadata and claiming that “nobody every reads metadata”. However, having lots of options for custom data sets in Digital Coast, particularly the Data Access Viewer (DAV), is now causing us problems with metadata clarity. How do I know people read metadata? Because they tell me when it’s wrong. Today I’m going to talk about how the metadata flows and which metadata to pay attention to.
Metadata on InPort
We’re now storing the Digital Coast metadata on the InPort system. The metadata that’s there reflects the data as it exists for custom processing in the DAV. For lidar point cloud data, that’s going to be geographic coordinates and ellipsoid heights. For imagery and DEMs it’s going to be whatever coordinate system the data originated in and, for DEMs, probably an orthometric vertical datum. This is the metadata you’re most likely to come across on the web. It reflects the starting point for anything you might ask to be done. It obviously can’t reflect the ending point because we don’t know what you’ll want yet. The metadata link in the DAV app is pointing to this version.
On the details page that you get by clicking a data set title in DAV, you’ll find a link to bulk download. For everything but lidar point clouds, this will be the same data files that are used for the DAV system and the metadata you’ll find there will be the same as it was on the InPort system. However, the point clouds are a little different. Few people want the point clouds in ellipsoid heights, so they’ve been transformed to orthometric heights. The index.html file will clearly state that and the metadata you find with the data will also show an orthometric datum and the processes used to get there.
If you do request some data through the DAV app, you are very likely to change some aspects about the data. That could be the projection, the file format, the horizontal or vertical datum, subsetting the data, or a number of other things, particularly for the point clouds. In that case, you’ll need to pay attention to the metadata file that comes in your zip file of data. This is typically an xml file, but there may be an html version of it too. We start from that metadata in the InPort system and change fields in the metadata to reflect what you asked for and add a process step to describe the changes. However, we don’t try to change things in the middle of a big block of text such as the abstract.
As with everything automated, things aren’t always perfect. The InPort system currently lacks a place to put the georeferencing. Apparently that made sense to someone. It’s getting fixed. In the meantime, we’ve tacked the original georeferencing information onto the supplemental information. Unfortunately, that spot is a little too prominent and it makes it look like that’s what the data is, even if you asked for and got something different through the DAV. We’re working on it and hope to fix some of that. In the meantime, keep reading that metadata!