EVS-Islands

Yes, I am cheap. If I can get something for free, that is my best price. It pains me to spend money, yet, I spent over $300 on my versions of Global Mapper. Now why would a tightwad like me spend money on something he could get free. (I am assuming that some of the open-source GIS packages could accomodate Landsat imagery and digitizing.) Many months ago, I downloaded the free trial version of Global Mapper and figured I give it a try. I was skeptical. I had my Landsat (MrSid format) and within seconds I had a 250MB file loaded and viewing it. It was cool! I was so impressed, I immediately got myself a working copy of the software and have been extemely pleased ever since.

The above slideshow details how one would load six Landsat images (about 1GB) in seconds. Within in one minute, you could be digitizing a shoreline of a target island. Global Mapper loads Landsat lightening fast. I know there are other packages that claim to be fast, but when I can load the image and begin digitizing within a minute, I'll spend my money with that company.

The pictures walk you through the steps to follow in order to load Landsat into Global Mapper. I was a bit careless when I first amassed my original collection of Landsat images and deleted all of the supporting files that allow for auto-georectifying. My extra steps involve, selecting the appropriate projection (UTM zone 28 for these images) and Datum (WGS-84 for Landsat ETM+).

Another plus is the excellent support a licensed user receives from Mike. Mike is Global Mapper and he personally makes sure everything works the way it is supposed to. Great service. Slick product, worth every cent!

Enjoy!

Labels: digitize, Global Mapper, Landsat

Posted by Mr Minton at 6/15/2007 03:11:00 PM | Permalink | Comments | links to this post

Digitizing from Landsat N-30-60_2000 ETM+ at 1:12,500

As my frequenct readers are aware, the 1:12,500 scale is what I typically use when I digitize EVS quality shorelines from Landsat ETM+ mosaics. I am able to distinguish features without having the image become a montage of different colored pixels. If I have a poor quality image, I might work at 1:25,000 to 1:50,000. A great image allows me to work at 1:6,750 occassionally. But typically, my work is done at the 1:12,500 scale.

I loaded this Landsat image into Global Mapper, zoomed in on an island shoreline, and began work. After three hours of digitizing, I decided to zoom out and look at the entire island group, which I assumed was the Faroes. They really looked strange!

Funny Looking Faroes?

I checked out my WVS map files and these Faroes were in the wrong place. They were right where the Shetlands should be. I checked out Wikipedia and located a MODIS image of the North Sea and there, just above the Orkneys, are the Shetlands and to the northwest of the Shetlands are the Faroes. They kind of look alike, but not really.

Orkneys, Shetlands and Faroes

Shetlands (left) and Faroes (right)

So, it looks like I am digitizing Shetland shorelines and not the Faroes. It seems that Landsat has dropped the ball on the Faroes. I am certain that they have imaged it numerous times, but none of these images are freely available to the public.

I'll be content to map the Shetlands for now and tackle the Faroes another day.

Enjoy!

Labels: EVS, Global Mapper, Landsat, MODIS, Orkneys, Shetlands, UK

Posted by Mr Minton at 6/14/2007 11:53:00 AM | Permalink | Comments | links to this post

"Does any GIS staff has experience with many types of visual analysis, can you tell me about some who can help us?" - Post from an Anonymous Indian Reader

I read this post to my wife. Upon completion, she said, "I don't get it! What are you trying to say?" "First the question doesn't make any sense and second your answer seems vague."

First - The question was posed by an English speaker and included poor grammar.

Second - More information from the commenter is required before a definitive answer can be given by the commentee.

Now, on with the original post. I personally think it reads well. At times, it tends toward the poetic, but basically it says "I need more information!". _______________________________________________

These are the words every blogger longs to hear, "...SOMEONE WHO CAN HELP US". And of course that someone who can help is you. Even though visual analysis sounds like something we all are quite proficient at...I mean, I study Google Earth's HiRes imagery. I can identify types of aircraft, types and sizes of automobiles and even landcover becomes a target for my keen visual skills. However, you professional visual analysts working at CIA, NSA and the Pentagon who locate potential targets in battlefields chuckle to yourself and say "This guy is in over his head. The skill required in visual analysis takes years to develop. Tish! Tish! Silly man!". Still others, typically connected with the "publish or perish" research community and using cool things like spectral analysis or Band 1-5 analysis do their own chuckle and say "This guy is in over his head. The software required to conduct serious visual analysis takes lots of dollars to purchase and years to master. Tish! Tish! Silly man!". And I say, "You are both right!" I cannot identify battlefield targets and do not have loads of dollars to purchase software or the years of expertise to accomplish "publish or perish" quality research. But you are wrong as well. So Tish! Tish! right back at you. All of us pondering this viewer's plea for help truely have no idea what they are after.

I researched the time frame of the comment and found the person asking the question is from New Delhi, India. Their command of English grammar is limited, but their plea for "help" was clear. The link they used to browse my site was a non-corporate type. If I had any SPAM flowing through my viens, I could get back to the person using that link - but I am SPAM free. I am left with a somewhat cryptic question "Does any GIS staff has experience with many types of visual anaylsis, can you tell me about some who can help us?".

The entrepreneur in me wants to go for it! (Tish! Tish! Silly man!). The blogger in me wants to go for it! (Tish! Tish! Silly man!). The sane, amateur map-maker who has lots of experience at digitizing shorelines of islands along with a specific subset of layers of information (vegetation, reefs, roads, etc.) cautiously asks "What is it you are trying to do? What is your budget? Explain your project goals. I will either be able to offer my limited expertise or I will direct you to groups that might be able to offer assistance."

Now, that is what I would say if that person were to visit my site, read this post and then restate their question in a more complete, informative way. It also says legions about who I have become after two+ years of doing what I do. (Tish! Tish! You wise man!)

Enjoy!

Posted by Mr Minton at 6/12/2007 06:53:00 PM | Permalink | Comments | links to this post

Let's begin this post by giving credit, where credit is due. My first attempt at shorelines for North Ronaldsay were extracted from imagery taken from MS Live Local. It appeared to be cloud-free Landsat imagery, but I'm not certain. The first attempt to map this island's shoreline resulted in very rough lines. These shorelines were "okay".

Intrepid searcher that I am, I began to scour the Internet for mapping of North Ronaldsay. The usual cast of characters showed up, but nothing great. That is until I discovered Get-A-Map, which uses Ordinance Survey raster map images. I believe the original maps were at 1:50,000 scale. I found another treasure trove of maps, text and images from a Scottish outfit called SCAPE. They do surveys of Scottish island shorelines to determine the identity and quality of archeological sites within 50-meters, or so, of the waterline. Their publications are available for free. Included is the Ordinance Survey maps (1:50,000) for the islands. In my case, all of the Orkney Islands, most especially North Ronaldsay.

My purpose is to produce the highest quality shoreline map of each island I study. In doing my work I pull imagery and data from a wide variety of public sources - Landsat, WVS, PGS, GNIS, and others. I always give credit to my sources and am keenly aware of copyrights. In this study I have extracted names of island settlements, features and hydro features from the Ordinance Survey 1:50,000. The shorelines were digitized from the same Ordinance Survey imagery used in the SCAPE study referenced earlier. With that all said, let's move on.

The first series of Marplot maps are derived from the MS Live Local imagery. As I stated, it gave me "okay" shorelines, but they were not up to EVS quality. All maps after the Get-A-Map image were derived from Ordinance Survey georectified imagery. Once loaded and georectified in Global Mapper, I digitized my new shorelines. They were great. They took the MS Live Local shorelines and finished them. I took most of my names for settlements, features and hydro features from the Ordinance Survey work. The 2nd series of maps also used a text entry feature within Marplot to pretty-up the labels. I really like the end product.

This project has taken about six to eight hours of work. I believe it is worth the effort. My only gripe concerns the text labels - they act like raster images when zooming. I can reselect them and shrink or expand them as needed, but that is a clumsy process when your map is loaded with labels. This is where the more complete mapping packages distance themselves from Marplot. Oh well, so goes life.

Enjoy!

Labels: EVS, Get-A-Map, Global Mapper, island, Landsat, Marplot, MS Live Local, Orkneys

Posted by Mr Minton at 6/10/2007 02:15:00 PM | Permalink | Comments | links to this post