EVS-Islands

According to USGS the Lambert Conic Conformal map projection is a good choice for small islands (large scale). The curious soul that I am, I wanted to look at a number of projections to see distortion in action. Because we take images from a globe and then try to flatten them, distortion is the foregone bane of map makers. Our goal is always to minimize distortion. I will use the projection recommended by USGS on future maps.

I put together a slideshow that displays a number of map projections using Lundy Island's shoreline. Click HERE, adjust the speed between slides to 1.5 seconds and enjoy the Map Projection slideshow.

OT: Comparison

Posted by Mr Minton at 5/13/2006 02:03:00 PM | Permalink | Comments | links to this post

An island located in the Bristol Channel. It has had human presence since prehistoric days. It's recent history is one of castles, Templars and churches. For such a small island, it is loaded with history. That is most probably the reason the island was purchased for the people of the United Kingdom by the United Trust. Using the Landsat image, some of the island features are difficult to distinguish. For example Rat Island appears to be connected to Lundy. However, in photographs from various web sites it shows as separate. The majority of the island's shoreline was relatively easy to digitize. The very fine details required a good deal of research before I finalized my map. The Lundy Island web site gives solid information about the island from a tourist's perspective. The Lundy Island Field Course is another interest web site with unique information about Lundy Island. I feel my map is a solid effort and can be integrated with other layers of information. OC: Atlantic

Posted by Mr Minton at 5/12/2006 10:32:00 PM | Permalink | Comments | links to this post

Back to Isle of Wight. I received an email that shared about an effort by a gentleman that had used a copyright free Ordinance Survey map (circa 1948) of the Isle of Wight to construct a shoreline map. His efforts were careful and cartographically solid.

So, the thorough map maker that I am, I compared my digitized coastline (using Landsat ETM+ N-30-50_2000) against the Ordinance Survey map file.

The Ordinance Survey shoreline is the green line and the Landsat ETM+ derived shoreline is red. Since the ETM+ image is georectified, I have great confidence that it tracks the shoreline as it was (in 2000). Whereas the Ordinance Survey tracks a 1940's best effort (most probably using aerial photography and land surveys).

Am I wrong or does the Enhanced Vector Shoreline (EVS) seem the most "true" representation of the actual shoreline?

OT: Comparison

Posted by Mr Minton at 5/11/2006 11:55:00 PM | Permalink | Comments | links to this post

My comments will be brief on this post. I spent about 4-hours to digitize, map and post the Isle of Wight project. I undertook this project after learning about this island from a visitor to this site. He shared about a project called OpenStreetMap.org. Members are encouraged to produce their own unique mapping. One of the members recently mapped the Isle of Wight using an old map (no copyright). I like my map of the island. It started out being a simple shoreline map. That changed when I worked in the rivers located on the island. I made some decisions about the ground cover in these river systems. Study the detailed map to see those choices mapped. I think the choices were good based on the imagery. I enjoyed working on this island. It has been awhile since I digitized a new area. Lots of fun. OC: Atlantic

Posted by Mr Minton at 5/10/2006 11:42:00 PM | Permalink | Comments | links to this post

If you visit my site with any sort of frequency you know my ambition is to map all of the shorelines of the world using Landsat ETM+ 2000 as base imagery. The scope of my potential endeavor is enormous for one person. The above map shows 622 Landsat mosaics (dark green and purple grids) that include shoreline imagery. I have all of them. Some of these areas are highly complex shorelines (southwestern end of Finland; Philippines; Japan between Honshu and Shikoku). Some of the areas (Greenland, northern Russia, northern Canada) are covered with snow effectively hiding the shorelines.

This quest to map world shorelines seems daunting, to say the least. Impossible, to be blunt. I have a vision of a world map with consistent, high-quality shorelines at scales between 1:125,000 and 1:62,750. I believe it will be accomplished using Landsat ETM+ 2000 imagery. However, to make it happen in my lifetime - I need help!

Doesn't this sound like an admirable quest? Like a "Mission Impossible". Your requirements are relatively simple - a computer, digitizing software that can handle MrSid formatted images (Global Mapper or others) and a willingness to carefully trace shorelines, following a cartographically sound protocol. Once you have completed your Landsat mosaic (1 of 622) you send the completed files to me and I, along with others, will stitch the files into a world shoreline map.

Honestly, it sounds so possible when I lay it out like this. All I need are willing digitizers and time. Any takers?

OT: Thinking

Posted by Mr Minton at 5/10/2006 07:10:00 AM | Permalink | Comments | links to this post

Just thought I'd give you all the full view of the entire island group. So far I've posted images and maps covering the southern and middle groups of islands. Still want to post a map and image for the northern group. If you want to explore this islands in greater detail, visit my post of the Peng-hu Tao (Southern Group). It is linked to a couple of informative web sites. OC: Pacific

Posted by Mr Minton at 5/09/2006 11:56:00 PM | Permalink | Comments | links to this post

This is the middle group of the Peng-hu Tao Islands. As I shared in an earlier post, these islands are off of the west coast of Taiwan. The Fisherman Islands are a favorite with tourists. The islands are low and reef fringed. The purpose of the above post is to show how the addition of all of the name labels can lead to a cluttered map. If you look at the map at this medium size, the labels appear to make the map to busy. However, if you look at the image at it's original size, the labels are less intrusive. OC: Pacific

Posted by Mr Minton at 5/09/2006 11:30:00 PM | Permalink | Comments | links to this post

FREE Global Orthorectified Landsat Data via FTP

I was working on a cloud covered island and figured there must be better imagery available. I am keenly aware that Landsat imagery is freely available. I typically work with Landsat ETM+ mosaics. The other way one can work with this imagery is using WRS-1 or WRS-2 sorted images. The trick is, you have to identify a Path and a Row - not necessarily the easiest task. Then I stumble across the above web site. You can snag all of the WRS images (circa 1970, 1990 and 2000). They also provide a link to a USGS Path/Row calculator. You enter a lat/lon and the applet calcultes the Path/Row. It is a slick interface. Once you know your areas Path/Row, then visit the ftp and download to your heart's content. I'll give you a step-by-step on how to get this free imagery.

(Some of you experienced "image hogs" might find the following obviously simple. I, on the other hand, think my audience appreciates a step-by-step guide. And since it is my site, I'll do what I want. Hope some of you find this helpful.)

---

This is the entry page for Landsat.org. It is here that you are linked to the path/row calculator and all of the image files.

This is the USGS path/row calculator. You enter your target areas lat/lon and the calculator determines the Landsat path/row of your images.

At the Landsat.org entry page select the range of paths your specific path is located in (WRS-2 Path 049-088).

Now you select your specific path (049).

Now you select your specific row (066).

This is your file (p049r066 7x010104).

If you want, select the jpg file and you can view a small picture of your image.

These are the files that you are able to select from. The images, except for the jpg, are tif files. They are compressed (zipped) and you will need to unzip them when you get them downloaded before using them. My digitizing/mapping software, Global Mapper, allows me to load zipped files. Once loaded, I can work them. Nice option. (I want to add - I know there are excellent mapping packages that match and often exceed the range of capabilities offered by GM. ESRI, the premier mapping package is one that comes to mind. However, GM offers a free trial and if you like it, it costs $199.00. That is big bucks for me. I am a software tightwad. I have not once regretted the $199.00 initial investment. I have purchased a major upgrade for $90. The sweet aspect about GM is it is simple, simple, simple to use. Loading map files or images is a snap. The tools are quick to learn and intuitive to use. But the absolute best feature is Mike. He is GM! Mike is programmer, problem solver and knowledge source extrodinaire. Mike makes GM a great investment. Whatever the problem, email him and he will respond within hours, often minutes. If the problem is in his software, he will fix it and send out a repaired version (no charge). For my bucks - GM is a solid package for digitizing from satellite imagery. 'nuf said. (No, I don't get a kick-back from GM - I just like it.)

This is a WRS-2 image, still zipped, about to be loaded into Global Mapper.

The image is loaded and I have selected my work area to zoom into.

Now I can select my tool to work with and begin digitizing.

OT: Data

Posted by Mr Minton at 5/07/2006 03:05:00 PM | Permalink | Comments | links to this post

As I digitized this shoreline, I came across what looked like a small bay. The WVS map files do not show a bay in this area. Without giving the feature too much thought, I digitized, mapped and finalized my product.

I looked at the finished map and pondered my decision. "If WVS doesn't show a bay, what could this 'bay' actually be?" The Landsat image is peppered with clouds. If you look closely at the map of Eiao, in the previous post, the green shorelines are classified as part of the EVS Shoreline Obscured layer. I studied the image closely and looked at the small cloud located SE of the bay. After determing that the shadows cast by all clouds display a distinct ground signatures - black shadow, which obscures all ground features beneath it and the shadow is cast in a NW direction from its originating cloud. I realized that the small bay was, most likely, a cloud shadow. I redigitized the shoreline and classified the reworked area as EVS Shoreline Obscured.

The redigitized shoreline is displayed above. Now, you ask, is this the actual shoreline? No, maybe, sure? In other words, I don't know yet. When I can study another image (Space Shuttle, map) of the island, I should be able to recast the shoreline into its proper layer (EVS Shoreline). Until that time, this cloud-shadow-obscured-shoreline is just that.

OT: How To

Posted by Mr Minton at 5/07/2006 05:10:00 AM | Permalink | Comments | links to this post

Eiao Island (8˚00'S., 140˚42'W.) lies SW of Hatutu; it rises to a height of 576m. The S coast of the island is formed by a perpendicular cliff, but the NW and W coasts are indented by several small bays. Its SE point displays a noticeable rock in the shape of an obelisk. The channel separating Eiao from Hatutu should be navigated with caution. A 6.4m shoal lies in the E entrance, while a reef, which has been reported to break, lies in the middle of the passage. The current generally sets NW through the channel with some strength.

The passage between Eiao and Motu Iti, further S, reportedly contains depths of 20 to 60m, and will be described with Nuku Hiva. Banc Hinakura, with a depth of 38m, lies 13 miles WNW of Eiao. A depth of 35m was reported 12 miles SW of the S extremity of Eiao, while a depth of 27m lies about 11 miles further SSW.

One can anchor at the opening of Vaituha Bay on the NW coast, in 27m of water, sand and mud, keeping the SW point of the bay bearing 247˚ at 0.2 mile.

In Charner Bay, also to the NW of the island, there is good anchorage, in 30m, sand and gravel, but it is not well sheltered from the swell which is felt strongly there. (SD PUB-126)

---

This is one of the first islands I became enamoured with. It is fairly large - 12kms by 4kms and 576m high. At one time, people lived on the island. Prior to European arrival, the island was vacated. It was visited by islanders to harvest stones to make adzes. The first Europeans to venture into the Marquesas were Spanish, then French and then American. The American, a Captain Porter, claimed to have discovered the northern islands (Nuku Hiva and north). He claimed these islands (including Eiao) as US possessions. Congress never acted upon his claim, so the islands remained French possessions.

At one time the island was used as a penal colony. Apparently, a few buildings/foundations still remain. In the 1970's/80's the French military suggested that the island be used to test nuclear weapons. They would drill into the island's core, drop bombs and explode them. The idea was rejected.

Today, the island is visited, on occassions, by divers. The island is infested with feral animals - goats, pigs, sheep, cattle and cats. No concerted effort to eradicate these island biota destroyers has yet been mounted. The feral animals are a continuing threat to the island's health. If you are a French authority reading this post, please sponser an effort to eradicate these feral animals from this small island.

Enjoy!

Labels: atoll, EVS, Global Mapper, island, Landsat, map, Marplot, Marquesas, Pacific, Polynesia

Posted by Mr Minton at 5/07/2006 04:44:00 AM | Permalink | Comments | links to this post