Thoughts - Ultima Thule, Gravel Bars and Curiosity

Peary Land, Greenland - The Image That Started My Search

On Sunday morning, while drinking my morning cup of coffee and checking my "usual websites", I came upon the above image of the northern shoreline of Greenland posted to NASA's Earth Observatory. I should have just kept right on moving, but I didn't. One image identified a point of land called Point Morris Jessup in the north center of the image. I decided that it would be loads of fun to check this part of the world out in more detail. About 8-hours worth of research, georectifying and jotting notes later and this is what I've learned . . .

Peary Land, Greenland - Kaffeklubben and Oodaag Islands

Kaffeklubben Island

The Ultima Thule is an island. Explorers from the ancient Greeks to those within the last few years, have searched for the northernmost land, Ultima Thule. Within the last 100 years, modern explorers think they have found it located off the northern coast of Greenland. In 1900 Peary sighted Kaffeklubben Island, but never landed. A Danish explorer, Lauge Koch, landed and named this lichen splattered spot of land after a coffee bar located in a Danish museum. The above aerial photo shows an oblique view of the island. It turns out the northern tip of the island is 750-meters farther north then Cape Morris Jessup, making it the Ultima Thule.

Oodaag Island - Ultima Thule?

That is until the above island, located 1360-meters north of Kaffeklubben Island, was located. Using aerial and satellite imagery, groups have pushed even farther north, by a few kilometers, finding more of these gravel bar islands.

(Before I move on I want to credit the source for the images, Arctic Thule. This site details the search for the northernmost land. It contains well written accounts of the latest efforts to identify the Ultima Thule.)

Here is my question - When is an island an island? Having mapped thousands over the past few years, I know that an island is a naturally formed area of land, surrounded by water, which is above water at high tide (per Article 121, Part 8). Based on that definition, Kaffeklubben Island is an island, but Oodaag Island is not as it comes and goes. Seems that once found in the 70's it was lost for a number of years and only recently rediscovered. It would seem that these gravel bar constituted islands to the north of Kaffeklubben Island are gravel bars first and, at best, temporary islands at least.

And finally, curiosity. I have spent way too much time on this topic. It is time to get ready for school and a classroom of middle school students. And I think I'm curious. You ought to spend time with young people, their curiosity is insatiable.

Enjoy!

Spain and Portugal - 1:100,000 Shoreline Image Tiles

Spain and Portugal Shorelines - 1:5,000,000 Landsat ETM+ Images (6-Images)

The above image of the Iberian Peninsula is a composite made up of six Landsat ETM+ mosaics. As an island person, I'm not much interested in most of this composition. In fact, I am only interested in the shorelines plus or minus 5 to 10 kilometers.

Spain and Portugal Shorelines - 1:100,000 J2 Tiles

So, I decided to create a shoreline image of the Iberian Peninsula. I first tried creating a series of 1:50,000 scale images of the coastline. It took way too many images and the scale did not include enought shoreline detail. I upped the scale to 1:100,000 and systematically created a series of 120 tiles to cover Spain and another 51 tiles to cover Portugal. Whereas the Landsat ETM+ images are 865 MBs in size, the 1:100,000 tiles for the entire Spanish and Portugese shorelines are 190 MBs in size (each tile about 1 MB).

N-31-40_2000 (100) ES - 010 J2 (1:400,000)

Let's zoom in and see what one of these shoreline tiles allows us to view. Keep in mind the resolution is 14.5 meters per pixel. So here it goes...

N-31-40_2000 (100) ES - 010 J2 (1:100,000)

N-31-40_2000 (100) ES - 010 J2 (1:50,000)

N-31-40_2000 (100) ES - 010 J2 (1-25,000)

How about that zoom capability? If the 1:100,000 image were of a shoreline I wanted to digitize, it would work just fine. I typically do most of my digitizing work at 1:25,000 to 1:12,500. These tiles would allow the end-user the ability to zoom into shorelines at sufficient scale to accomplish EVS precision shoreline mapping.

If you are interested in Spanish or Portugese shoreline mapping work, these 1:100,000 scale image tiles would do the trick.

Enjoy!

Melilla Border Fence Limits Study

Melilla Enclave

I enjoy enclaves. The Portugese scattered these trading ports along the costs of Africa and India. The Spanish were more into settling, however in the case of Morocco, they established enclaves - Ceuta and Melilla. Above is an aerial photo of the limits of Melilla looking from the port toward the northwest.

Melilla Enclave Border Fence Limits Study - Landsat ETM+ N-30-35_2000 (1:40,000)

Melilla Enclave Border Fence Limits Study - DigitalGlobe Image from Google Earth (1:40,000)

Using the above two base image sources, I created a shoreline map of Melilla. Working first off of Landsat, I created a rough EVS precision shoreline map. Working with DigitalGlobe imagery from Google Earth, I decided to not only track the shoreline using 1-meter resolution imagery, but also to determine the course of Melilla's border fence.

Melilla Enclave Border Fence Limits Study - DigitalGlobe Image from Google Earth - The Border Fence

Melilla Enclave Border Fence Limits Study - BBC Image_40847408_twosides220afp

Back in 1977, I travelled to Berlin on business. While there, I visited Checkpoint Charlie and viewed the Berlin Wall. I hated that wall! It stopped what should have been normal flow of peoples and goods within Berlin. I get the same feelings when I look at the Melilla Border Fence. It was built to keep Africans out. Not only out of Melilla, but ultimately out of Europe. It seems the "Haves" cannot accomodate the "Have Nots". The world just has too many hungry people for all of us to care for. I'm sorry. I'll get off of my soapbox and get back to my map.

Melilla Enclave Border Fence Limits Study - EEVS Precision Map (1:40,000)

My international borders come from the NGDC Coastline Extractor. Specifically from the dataset called WDBII International Borders. These borders are used in many global mapping packages. They work, kind of. In my above map I traced the Melilla Enclave Border Fence from start to finish. I'm not sure if that represents the actual border between Melilla and Morocco, but it seems pretty final to me. My Melilla map shows the WDBII borders (orange line) and the fence line (green line).

So, if you want to create the definitive Melilla - Morocco border, I'd say the fence line seems to be a natural dividing line. I wait for the day border fences that separate people unnecessarily can be removed and allow us to move freely as we wish and as we should.

Enjoy!

Musandam Peninsula OM - The Peninsula of Peninsulas

The Strait of Hormuz - US Government Map

Strait of Hormuz - Landsat ETM+ N-40-25_2000 (1-1,000,000)

Let's begin this post with an image of a US Government map of the Strait of Hormuz, followed by a Landsat ETM+ image of the same area. Iran is to the north and Oman and the UAE are located to the south. Extending into the Strait of Hormuz from Oman is a rather odd shaped peninsula. It is called the Musandam Peninsula. It is the subject of my latest EVS Precision mapping efforts.

I teach early Greek history to middle school students. Every year I show them the Greek landform I call the "Peninsula of Peninsulas". You and I know it as the Peloponesian Peninsula. It contains four distinct peninsulas. Now let's take a look at the Musandam Peninsula.

Musandam Peninsula OM - Landsat ETM+ N-40-25_2000 (1-175,000)

Musandam Peninsula - EVS Precision Map (1-175,000)

My conservative count of the peninsulas within this peninsula is 12 to 13. The shoreline of the map does not do the complexity of this peninsula justice. I mapped the entire shoreline of this portion of Oman at 1:12,500 scale. To appreciate the shoreline complexity, one should zoom into this twisted piece of real estate using Landsat ETM+ imagery. Google Earth and Google Maps contain the Landsat imagery for this area. The digitizing took in excess of 12-hours to complete.

Located on the northern coast of the Musandam Peninsula is a city of 3,000 hardy souls. The small city of Kumzar, where people speak Kumzari, was isolated from the rest of Oman until recently. The people fished and stayed connected with the other Omanis by sea. Now days, they are able to use a helicopter to fly to and from the larger Omani cities. NPR has an interesting news piece that shows pictures and contains first hand information about Kumzar. Apparently the area offshore of Kumzar is highly regarded as a dive site. This website has a number of shoreline and underwater photos of the area around Kumzar.

During the brutally hot summer months, the Kumzarians, leave their northern home and travel to Khasab which has more tolerable summer weather. After summer has gone and milder weather returns, they go back to their isolated northern village.

Enjoy!

Anaa Atoll FP - Many Motus Surrounding a Shallow Lagoon

Anaa Atoll FP - Landsat ETM+ Image S-06-15_2000 (1:125,000)

Anaa Atoll FP - ISS006-E-34723 and 34724 Image (1:125,000)

Prior to beginning an EVS precision island mapping project, I scout for ISS imagery, especially ISS-006 imagery. Some excellent cloud-free images were captured during this mission. In the case of Anaa Atoll, Landsat ETM+ had some cloud issues. I downloaded, combined and georectified ISS imagery. Working with the Landsat image beneath the ISS image gave me excellent detail from which to create my finished EVS precision map.

Anaa Atoll FP - EVS Precision Map (1:125,000)

For me, the most interesting feature on this map is the shallow lagoon. It has a few deep areas, but is overall shallow with most underwater features within it close to the surface. The atoll has a permanent population of 435 people (2002 census) scattered between a main village and a couple of smaller settlements. According to the Wikipedia article about Anaa Atoll, the soldiers that lived on Anaa were an especially viscious group. In 1769, James Cook was the first European to sight the atoll.

This mapping project took about three weeks to complete using an on-again, off-again pace. Seems my life has gotten complicated between preparing for the new school year, watching my grandchildren, completing "honey-do's" and squeezing in map work late at night.

Enjoy!

Comparison - MrSID, JPG, JPG2000, GeoTIFF 24-bit and GeoTIFF 8-bit

There are times I wish to work with Landsat ETM+ imagery with attention to file size. Instead of 80 MBs of raster imagery, I would like to work with 4 MBs. Using Global Mapper Raster Image Export feature, I am able to get the imagery I need easily. The raster format I typically use is JPG2000. Let me share why.

Raster Comparison (1:300,000)

I did a comparison of different raster formats July 2007. My previous comparison was poorly constructed and explained. I always meant to redo it. So here it goes. The first composite shows a portion of a MrSID, JPG, JPG200, GeoTIFF 24-bit and GeoTIFF 8-bit. At 1:300,000 scale the images are visibly indistinguishable. One appears to be just as good as the other.

Raster Comparison (1:25,000)

At 1:25,000 scale, which is pushing the limits of the original image (MrSID format), the first 4 images are very similar in clarity. The GeoTIFF 8-bit image begins to lose color depth, as one would expect from 8-bit imagery as it uses fewer colors in its color palette.

Raster Comparison (1:8,000)

Raster Comparison (1:2,000)

Finally, lets push the images far beyond their intended usefulness. The three middle images, JPG, JPG2000 and GeoTIFF 24-bit are noticeably superior to MrSID and GeoTIFF 8-bit at this scale. It is at the 1:2,000 scale that pixel size difference shows up. The three middle images all carry the same sized pixels and range of colors. The GeoTIFF 8-bit image has the same pixel size as the middle images, but without the rich variety of colors. The MrSID image, our base image, has a larger pixel size. Its color palette is used to form the colors found in all of the other images.

Now for the kicker - size and number of files: MrSID - 80.4 MBs and 2 files; JPG - 21.5 MBs and 3 files; JPG2000 - 4.5 MBs and 3 files; GeoTIFF 24-bits - 121.8 MBs and 1 file; GeoTIFF 8-bit - 10.4 MBs and 1 file. Clearly, the JPG2000 packs the greatest image bang for your MBs. However, I understand that JPG2000 is not widely used. If you are able to use it, do so as it gives you quality imagery in a compressed package. If you are not able to use JPG2000 format and you do not require working at 1:50,000 or lower, the GeoTIFF 8-bit works quite well. I happen to prefer JPG as my second choice after JPG2000. At twice the size of GeoTIFF 8-bit imagery, it is still a reasonably sized file to work with in lieu of MrSID. GeoTIFF 24-bits is far too large compared to the other three raster formats. Last, but not least is the MrSID format. It is my base imagery. The four other raster images were derived from my Landsat ETM+ image in MrSID format.

Hopefully, this raster comparison is both informative and helpful. If nothing else, the raster comparison images are pretty slick composites giving you side-by-side visible examples of these 5 popular raster formats.

Enjoy!

Kasatochi Island - Site of an August 2008 Volcanic Eruption

Kasatochi Island - Landsat Image from N-01-50_2000 (1:16,820)

Kasatochi Island - ASTER-Terra Image AST_2003273 (1:16,820)

>

Kasatochi Island - EVS Precision Map (1:16,820)

I am an opportunistic island mapper. When I encounter an interesting island and it is mappable, I map it. So it goes with Kasatochi Island. Today's article in Earth Observatory Newsroom detailed the recent activities of Kasatochi Island. Located toward the western end of the Aleutians, this typically dormant volcanic island erupted on August 7, 2008. The eruption spewed out such a large amount of ash that Alaska Airlines cancelled 44 flights. The volcanic activity has continued and is being monitored daily by the Alaska Volcano Observatory (AVO). Follow the aforementioned link and it will give you all of the latest facts concerning this island's volcanic activity.

The Aster-Terra image is dated 09/23/2003, well before the current eruption. The Landsat image predates this by at least two years. I have no post eruption images, so all you get is a before map. The island is covered by vegetation, most probably low lying grasses. I am certain this eruption has refashioned the island's ground cover. At the bottom of the caldera was a small lake. It was probably vaporized in the eruption. When post eruption images become available, I will make a post eruption map to go with the above map.

Enjoy!