Down to the Wire

It's the last week of classes - then the final presentation. I have lots of writing to do on my paper, and lots of organizing files for Dr. Poudel.

So much work, so little time. However, some of the outstanding questions about the NEDs I was working with have been answered.

2009 Imagery!

I received the 2009 imagery from the Alexandria, LA NRCS office on Monday, October 26 and have been working with it ever since. The imagery includes all of Vermilion Parish, so I've trimmed it down to the Coulee Baton watershed, including the downstream part of Coulee Baton that is not part of the microwatershed project.

The land use polygons have to be created manually by hand, and I've started on the human habitation polygons. I still need to touch base with the Abbeville Farm Service office for current crop data.

I have the 2006 maps on one screen, and the 2009 imagery on the other - the apparent human habitation is much more dense on the 2009 than on the 2006.

Beautiful Images

It's easy to lose time staring at the beautiful images created by the ancient meanders - wish I knew more about the geomorphology. I only remember the basics of it from historical geology.

More Watershed Analysis

I have downloaded 3-meter resolution NED files from the USDA Geospatial Data Gateway. I have mosaicked the ones from my study area, and extracted a smaller area of the mosaic using a mask.

I have run the hydrology tools on this NED, and have had much better results than with the first DEMs (10 meter resolution). I want to run them using pour points instead of the area-wide tools I've been using, see how those results compare with my current results.

My main problem is, the textbook I'm using doesn't go into as much depth as I'd like on this subject. It references other works, but so far I haven't gotten my hands on them. I have questions regarding the accuracy of the watershed analysis tools. My watersheds almost, but not quite, agree with the hydrology unit layer I've downloaded from the USDA website. I've downloaded both the 8- and 12- digit layers. What do GIS analysts do when confronted with this kind of conflict?

I've been spending all my time on the watershed analysis tools, and I need to get busy with the basics - land use, etc.

Bob Capezza is supposed to send me the 2009 NAIP files for Vermilion Parish via DVD. I've downloaded the 2007 layer, which is the latest available on the USDA web site.

I need to contact the Abbeville NRCS and Farm Service people on the land use data because previous GIS information on Coulee Baton was lost in a computer crash.

Not Thrilled with Results of Watershed Tool

I ran some tools on my mosaicked DEM, got some good preliminary results, but the result of the watershed tool was a value of 1, which gives a map that is one solid color. That's not what you want.

I need to re-run some of the preparatory tools again with different values, see if I come up with any new results. I might not, because the D8 algorithm in ArcMap is fast and efficient of computer resources, but is known to work much better on areas with more relief than I have in my very flat watershed.

Coulee Baton DEMs

I've downloaded the LiDAR DEMs and contours for the study area. I have mosaicked the DEMs into one raster preparatory to running the ArcTools watershed analysis tools.

Next step - contact the NRCS people involved in the project planning and see what kind of land use or other data layers they might have that will be pertinent to the project.

Field Trip to Coulee Baton

I'm looking forward to my field trip this morning to the study area of this semester's GIS project - Coulee Baton. It's north of Kaplan and Abbeville. I have been downloading the DEMs and contours for the area, trying to get them to look right because the study area comprises 4 quads.

Still studying up on terrain and watershed analysis possibilities in ArcMap. As suspected, those algorithms work much better on very hilly terrain than the flatness of my study area.

Next Thing

Jim would like an aerial photo map of the Bayou Sorrel indian mound. He showed me where it was on one of his maps; now I need to find the right DOQQ.

We'll see!

Getting It Together

Since the husband flew off to Canada last weekend, I've been trying to get back to working on my "summer school" project. It is an expansion of last spring's class project.

Got everything going, overcame some squirrely ArcView issues, and it's coming along.

I've got some modern DOQQs of the area, some scanned 1:250,000 topo maps (which unfortunately split the study area horizontally right about in the middle), and some historic 15 minute quads. Got them clipped down; they fit together pretty well. However, the colors and symbology is not consistent across the quads so they don't look very smooth.

We'll see what Jim needs to show on his Powerpoint later this afternoon, decide what will fit.

Oops - Can't Work on GIS Today

Got a call this morning to go into work - I will probably have another day off later this week to catch up on my GIS project. Can't do it today till later.

Need To Get Back to Work

I have taken a few days off this project; I need to get the next phase completed. Tomorrow and Monday should give me a chance to make some progress.

Moving Right Along

I have clipped and combined the 9 quads that are needed to show Jim's study area. Next step, create the line and point layers.

The Work Continues

Jim Delahoussaye is giving a presentation at Vermilionville for their lecture series in mid-July. I'm working on expanding my class project to give him some maps he can use in the slide show about the Myette Point people.

I don't know why I couldn't figure out this feature of ArcMap last semester; I guess I was so panicked by the end of the semester that I wasn't paying attention.

Anyway, I had previously used clipping in last fall's class, and while revisiting it a few days ago, I found that it is possible to clip a base map to the extent of a layer. This will effectively clip the map collars of the historic quads I'm using, and provide a nearly mosaicked map. I think I have all the quads of the study area in the 1935 edition, so I'm getting excited about working on it.

How'd I do?

Just for the record, I got an A in the GIS 465 class. Nice thing to match my A in GIS 455 last fall. It's been 30 years since I graduated from LSU so going back to school was a stretch in some ways.

The teenager in the house would never have let me forget it if I had made anything else.

Final Project Presentation Done

Monday night was the final presentation for work in GIS 465. Jim Delahoussaye and Ray came, were very kind in their comments. I enjoyed the other students' presentations, and hope mine went OK. Hard to tell while it's going on.

Got some good feedback and ideas on my project from fellow students and instructors. Grayden suggesting using demographic data in my town table, using that to make the symbology change with population size. I can add that easily. Also, I think my landmark table needs revising; Jim intimated that some corrections were needed there.

Brent suggested using historical aerial photography imagery; might exist dating from 1970s.

I would like to put the sandbar development in a Flash presentation, and also mosaic all the 15 minute quads to encompass the whole study area, not just the Foster quad.

Now, on to find 2008 LIDAR data for Grand Isle.

Grand Lake Sandbar Deposition 1935-1973

Grand Lake Sandbar 1973

Grand Lake Sandbar 1959

Grand Lake Sandbar Formation 1948

1935 Map Grand Lake, Louisiana

Results of Polygon Creation Efforts

Here is a draft; I'm still working on the polygons, the shorelines, and need to put map information text. Still very unfinished but it might work.

Still Creating Polygons

My polygon layers represent the growth of the sandbars in Grand Lake between 1935 and 1973. I have over 500 objects in three layers. I'm about to start on another sandbar; the largest. It should be between 250 and 300 objects.

Also have point layers showing individual locations at different points in time, and some shoreline layers.

Nothing Like Digitizing First Thing in the Morning

Added a few polygons to my sandbar layer while drinking coffee. Now I have to hurry and get ready to go to the office.

Tried to interest the teenage daughter in the project last night; thought she could get interested in doing some polygons. Attempted a Tom Sawyer-like psychology maneuver but daughter saw through it, pronounced the polygon digitizing activity "tedious."

Well, it can be. I choose to think of it in a more zen-like manner, and absorb the beauty of the landforms. While I try to suppress the muscle spasms in my neck.

The End is in Sight

Our final exam is scheduled for 5 pm Monday, May 4. We will need to present our project in PowerPoint to the class, and turn in a written report and CD with our work on it.

I finally have georeferenced historic maps, and I'm concentrating on the Foster/Centerville 15 minute quad. I'm busy making polygon layers showing the sandbar growth. The shorelines are changing too; so I am going to try to do a shoreline layer representing 1935, 1940, etc.

The geomorphology is more interesting to me at this point than the cultural emphasis that I started with. After tracing so many lobes and mittens of the delta, I'm beginning to be fascinated by the stages of its formation. I am doing a new layer for each point in time, which will have the additional feature of being able to approximate the area of the delta in each time period.

Trying to Make Progress on Final Project

Now I have the maps from DEQ - about 90 of them. I separated the 15 minute and 7.5 minute ones in separate folders. Looks like I won't be able to trim the map collars and make them stay georeferenced. They don't mosaic well but I need to start creating my feature classes anyway.

I planned to spend the whole weekend working on it but other matters have come up. Hope to have some time this evening. The original project will not be completed, but I hope to show progress. Jim Delahoussaye is still working on the book that my project will illustrate/complement.

I've about decided to complete showing the sand bar growth in Grand Lake as the major part that will be completed during this class.

One of the presenters at the conference said, in passing, that she had a set of maps for Grand Lake so I need to find her e-mail, see what she has. Can't wait for it though, have to soldier on with what I have now.

25th Annual Remote Sensing & GIS Conference

The conference kicked off this morning in the new conference center on LSU campus. Beautiful space, very nice people. Heard some very interesting presentations this morning. I'm going to use a few phrases I heard in my final presentation. You ready? Here they are... "hydrologic regime" and "multi-temporal analysis."

I got to meet Joe Holmes, newly retired from DEQ, who was the source of maps for my project this semester. Met others from the USGS in Lafayette, also people from Lincoln Parish.

Time to get back... lunch about over.

Georeferencing workshop

I georeferenced a street map of LSU to a raster screen shot mosaicked in Photoshop. I got RMS down to 1.8. Very nice map.

Very good presentation, learned a lot. It's possible to use building layer as well as street layer to georeference. Now I know what I was missing by relying on the ArcMap help screens.

Monday morning in Baton Rouge

Drove through the fog this morning to attend a georectification workshop preliminary to the LARSGIS conference this week. Waiting for latecomers to arrive. Hope to gain mental illumination on the topics I've been struggling with this semester for my class project using historic maps.

Maps Galore

Jim brought by the external hard drive that he copied 90 area maps from DEQ. I can't wait to have time to look them over.

New Stuff!

I got some new tiffs today from Joe Holmes at DEQ - they are georeferenced historic 15 minute quads in my study area. They are referenced to NAD83. I haven't tried to create my feature layers yet with them, but I downloaded, unzipped and looked at them in ArcMap. I am regaining hope for the project. Meanwhile, during lab yesterday, I downloaded some modern maps that I was going to try to create polygon layers showing the shoreline changes and sandbar development.

I'm going to go forward with the maps Joe sent me. My cohort in crime, Jim Delahoussaye, will go to Joe's office tomorrow with an external hard drive to copy all the historic maps that Joe has on the Atchafalaya Basin area.

Answers on the Horizon

I met John Snead, Cartographic Manager at the Louisiana Geological Survey at the cochon de lait we attended today. I mentioned my problems with the 15 minute historic quads, and he told me that the answer lies in the properties of the projection. Specifically, the meridian, lat/lon and other properties. I need my projection to match.

He suggested that there may be metadata available on the maps, which would be contain the information I'm looking for.

I'm going to e-mail Joe Holmes at DEQ to see if metadata exists on these maps.

All Set for Rectifying Rasters Pre-Conf Session

I have signed up for the pre-conference session on rectifying rasters. It is only a morning session, but I hope to at least find out why my 15-minute quads won't work.

GIS Woes Continue

This is not an April Fools joke - I have to figure out how to complete the project without my favorite maps. Time's a-wastin'. I also need to finish all the reading to find the locations for my features.

Why do the 7.5 minute old maps georeference fine, but not the 15-minute ones? Still reading up on ESRI web site support and knowledgebase as much as time allows; haven't found answer. Bryan, the lab instructor, has my files and is giving it a try.

Looking forward to the Remote Sensing and GIS workshop in Baton Rouge the week after Easter.

ArcGIS 9.3

I received my student edition of ArcGIS 9.3 last week, got it all installed. I was hoping the new version would make some difference with my troublesome Foster quads. However, I had the same results when attempting to georeference. The Foster quads want to be 1/2 degree off, even when the map frame has the correct projection.

My hope of mosaicking the historic quads is over; now I have to figure out how to tell the story with the independent quads.

It's wonderful to have the software at home. I wish I had dual monitors like the lab because it's also handy to have ArcCatalog open in one monitor and ArcMap open in another.

On the Verge of an Epic Fail

This may be an exaggeration, but I'm pretty bummed right now. The semester is more than half over and my favorite base maps are still misbehaving.

I can georeference the other historic 15 minute maps, but not the Foster quad. I think it's a projection error. However, the map collar info on the Foster quads looks the same as the Chicot and Loreauville quads that I've successfully georeferenced. But none of the projections that work with Chicot and Loreauville will work with Foster. The Foster quads are all 1/2 degree off, and squish vertically when 3 coordinates are entered with the georeferencing tool.

The Foster quad shows the sand bar growth over time in Grand Lake, and is the most dramatic of the maps.

I have ordered a student version of the software so I can work at home; I need to be able to work on this on weekends, not just in available lab hours.

How I've Done So Far

Ok. I have done two project proposals this semester. The first one fell through due to lack of geographic data. It might work in another year if the samples are identified geographically.

I have another one, less spatial analysis possible but more descriptive. I have many text files to read to gather the spatial data from. This project has potential to be a cultural history map.

I have scanned 15 minute maps, from the mid-1930s to the mid-1970s. I might be missing one or two from the study area, need to see about getting them.

I was successful in georeferencing one of the maps two weeks ago at the end of lab time, thought I was home free. Worked last lab at georeferencing the rest, but it didn't work. They were all half a degree off; when I put in correct coordinates the map would distort, squished together sideways. It looked OK if I put the wrong coordinates. I fooled around with the projection properties in the data frame because the distortion looked like a projection error to me. Could not find a projection that would remove the distortion.

So far, have not found a reason for this happening, have not found any more information on possible reason/solution on ESRI web site yet. Need to find a forum.

Other steps - need to contact DEQ source for historic maps, see if I can copy the Atchafalaya maps onto a portable hard drive. A little bird told me source is going to retire this spring, need to move fast.

Why???

Because I'd like a record of my progress. I'm taking the second GIS class in a series, and running up against some roadblocks.

I'm creating this as a place to vent, and also to record my triumphs.