Comparing pictures- Find the differences- Lab 4

Comparing pictures to find the items and missing or changed things has been one of my favorite pastimes.  This exercise was an extensive grownup version of a familiar child's activity.   There's an amazing science which has developed to statistically calculate the accuracy of the human verses the computerized (production) method of classification.   We take an undated aerial of Pascagoula, MS and Moss Point, MS (next to one another) and verify our original land use/land cover classification.  Here we had to split hairs so to speak so if the point did not land exactly on something the spot was invalid.  We did not calculate a "within radius" of any intended target which may have raised the accuracy some.   Anyhow, I was 9/30  at 30 percent, not so bad for the first attempt.  My truthing points were truly placed as quickly and as randomly as possible to keep from having a "selected spot" type bias which likely accounts for many of them landing "out there".   Here's the map and my table of truthing with notes.  

It's all about the Prep- Getting ready for the Flood

We are about to engage in the  Network Analyst extension.   Here's a prep map for the new layers we'll be using to create escape routes for a portion of the Tampa Bay area in case of a surge.. We set up the layers using two python scripts, one for mass clipping and a second for mass projection.  These were great to use. They save a lot of time and keep the errors down.  We reclassed the DEM layer then converted it to a polygon.  I went ahead and separated out the elevation values so I could see the range on the map.  Pretty cool, so much of this area is under 20 feet.  If the recent storm had not calmed before it hit during the convention more of the area would have been under water.  Our flood zone was for areas under 6 feet.  It is surprising to see how large an area this covered.

  This was a real scenario just a few months ago.   Did they use this to prepare?  

Looking at Pictures can tell a Story- From Above

Oh My!  It is just amazing what can be identified from an aerial photo.  This one of Pascagoula, Mississippi shows so much.  One can see schools, cars, lakes, businesses.  We have been learning how to identify land uses by reviewing the associations, shapes, sizes and such of structures.  Here's our first land use map, there may be many more to come in this class.

Here's a peek at the attribute table for the commercial category.  There was not enough room on the map to set out the full listing of categories this time.