User talk:JimIrwin

From Wikipedia, the free encyclopedia

Contents 1 Great job! 2 Map data? 3 Thanks for the maps! 4 CDP maps in New Jersey 5 Coordination and questions re CDPs 6 Matching CDP and Township names 7 Index map captions 8 Highland Lake vs. Highland Lakes 9 Maps 10 ZIP Codes and ZCTAs 11 Question about your maps for counties in New Jersey 12 Oh, nice job! 13 Redundant Images 14 Really? 15 Mislabled highway on map of Hawthorne, NJ 16 Hello Jim 17 Railway map? 18 Maps Query 19 Paulins Kill 20 New SVG maps 21 Paulins Kill now a Featured Article 22 Population maps 23 New Hampshire's population density map has an error

[edit] Great job!

You did a great job on all your maps. I salute you! Speaking of which, would you happen to have a map of Monmouth County with all township boarders? I'm a resident of Marlboro Township. Thanks.

Monmouth County may be the one county with the most number of small incorporated municipalities, and it makes it difficult to show all the legal borders. The townships could be shown easily, but there are 20 or more small boroughs along the coast that hardly show at all. I could add an outline map to the county page, if it would be helpful, but it would be hard to label everything. -- JimIrwin 12:28, 22 December 2005 (UTC)

I added an index map to the Monmouth County, New Jersey page. Take a look and see what you think. --JimIrwin 14:10, 22 December 2005 (UTC)

Feedback: I think it's great! Keep up the good work. - Chris

[edit] Map data?

Hi Jim, I saw your recent uploads of Image:Andover boro nj.jpg and Image:Andover twp nj.jpg and I was wondering where the map data came from, and what tools you used to put it together. I'd like to make similar maps for other areas. Thanks. --ChrisRuvolo (t) 01:59, 2 December 2005 (UTC)

The software was Manifold, the data was from multiple sources. The administrative boundaries and water features were from the Census Bureau's 2004 Tiger/Line (second edition) files for Sussex County, the road network was from the New Jersey Department of Transportation, and the place names and locations were from the USGS Geographic Name Information Service files. I used information from the Tiger/Line files to identify the divided highways. The New Jersey state outline was from the Bureau of Transportation Statistics 2004 National Transporation Atlas Databases state boundary file. --JimIrwin 03:35, 2 December 2005 (UTC)

Interesting, but I think Manifold may be out of my price range. Thanks for your contribtions, these are quite complex maps compiled from several public domain sources, good work. BTW, is it possible to upload PNG or SVG output from Manifest? The current JPEGs have compression artifacts, as JPEG was intended for photos, not solid-color-region graphical data. See Wikipedia:Preparing images for upload. Thanks. --ChrisRuvolo (t) 16:53, 4 December 2005 (UTC)

Thanks, I changed all the maps to PNG files. -- JimIrwin 00:24, 6 December 2005 (UTC)

[edit] Thanks for the maps!

Hi Jim, I also appreciated your maps for Sussex County and the ones that seem to be appearing as I type for Passaic County. I have been part of a small group working on Wikiproject New Jersey, a project aimed at filling in the holes and expanding the details regarding New Jersey in all its aspects. While there are portions of the state covered by your maps, and those ChrisRuvolo has created, we could definitely use maps for all of the 566 municipalities in the state (plus the CDPs and unincorporated areas if possible). Any assistance that you could provide in creating and uploading maps, or in providing guidance in their creation, would be most appreciated. We could divvy up the map creation work so that we can ensure that the whole state is covered, and not just by one or two people. Alansohn 21:21, 5 December 2005 (UTC)

I think the maps are great, and I see that Warren County is underway. My only question is that on some of the maps, particularly for smaller municipalities (e.g., Belvidere, New Jersey), it can be hard to distinguish between the highlighted municipality and the rest of the county. Is there any way to change the colors to improve the contrast? Other than that, keep up the great work! Alansohn 01:26, 7 December 2005 (UTC)

Some of the smallest places do get lost. The larger red label was an attempt to highlight the principle location on each map, but at the scale of the map, there isn't much to be done for the small places. I started out mostly interested in showing the township outlines, since most of the online street map servers don't show township outlines. Once I have a county set up, it is relatively easy to produce an entire set of maps all at the same scale, but it would be much more work to change the scale or colors for individual towns. Maybe there will be time to revisit the smaller places later on. -- JimIrwin 01:52, 7 December 2005 (UTC)

[edit] CDP maps in New Jersey

I wanted to thank you for your work on maps of New Jersey's census-designated places. I had been searching for weeks to find a source that explains where the CDPs are, and if they are coextensive with a muncipality, within a municiplaity or in two municipalities. I had just found the data on the United States Census Bureau web site, and i will be following up to update county, CDP and municipality pages (as you and I have both been doing) and ensure that all CDPs are properly referenced. thanks for picking up my Lake Mohawk, New Jersey mistake. Alansohn 00:53, 18 December 2005 (UTC)

You're welcome. I intend to create maps for all the NJ CDPs. According to the census bureau documentation, a CDP will never overlap any portion of an incorporated city. Nevertheless, the CDPs are defined at the time of a decennial census, and don't change for 10 years, but the borders of legal entities can change, so it is possible, if unlikely, that a CDP or portion of one will become incorporated into a legal municipality sometime over the 10 year period. The data I've used for the borders of municipalities was accurate as of August 2005, and the borders of the CDPs were from the 2000 census. JimIrwin 10:38, 18 December 2005 (UTC)

[edit] Coordination and questions re CDPs

1) You seem to be a step ahead of me in associating CDPs with townships on the county pages. I am going through each county, in alpha order, and updating the pages for each CDP to list the parent munipality (or municipalities, as may be the case), and for each parent township to list the CDP (or CDPs or portions thereof) contained within. 2) How should we deal with trivial CDPs, where the Township and CDP are the same. In almost all cases where the CDP and township name match, they are one and the same. Some of these (e.g. East Brunswick, New Jersey) have pages for both and are listed on the County page, but most aren't. Should we even bother to keep the trivial CDPs where they exist? In some cases, only the Township page exists, but it indicates that it's both a Township and a CDP. Do we need to list the CDP status? But note that Cranbury Township, New Jersey and Cranbury, New Jersey are not the same. 3) I will be creating a List of census-designated places in New Jersey that would show all of the non-trivial CDPs, perhaps grouped by county within Township. 4) There is an entry for a CDP Titusville, New Jersey in Mercer County, New Jersey under Hopewell Township, New Jersey that doesn't seem to exist on the Census Bureau website? Any thoughts? --- Thanks for your work and any assistance or suggestions that you can provide Alansohn 22:20, 21 December 2005 (UTC)

Sorry, I wasn't trying to race you, I was just interested in getting maps of the CDPs uploaded. For the ones that were exactly the same as the township, I didn't bother with a map. I think I have maps for all the nontrivial CDPs now. If you come across one that doesn't have a map, let me know, because it would be an oversight on my part. My opinion is that the trivial CDPs probably shouldn't have their own page, since their census statistics will be the same, and there is no history, etc. that could possibly be different.

In principle, over time it is possible that a township could have a border change and the CDP would then be different from the township, but that is a remote possibility and hardly a good reason to double list everything. I think keeping the township page and making mention that it is also a CDP is a good solution.

I don't see the Titusville CDP in the Census Bureau's Tiger/Line 2004se files, nor is it in the geographic headers of the Census SF1 dataset. When I look at the page history, I don't see anything earlier than February 2005, which seems suspicious. I wonder if someone just copied some other page's stats and created a bogus page? The only place I see any mention of Titusville is as a USGS GNIS place name. -- JimIrwin 00:37, 22 December 2005 (UTC)

There's no reason to apologize. Whoever does the updates first doesn't matter. Other than that, I'll proceed as discussed. I'll check further into Titusville... Alansohn 01:07, 22 December 2005 (UTC)

I have a list from the State of New Jersey that shows CDP Holiday City South, New Jersey split between Berkeley Township, New Jersey (the only place where you have it) and Dover Township, New Jersey. Are you sure that it's not the case? Any thoughts? Alansohn 05:06, 23 December 2005 (UTC)

I reviewed the Census Bureau's Tiger/Line borders, and Holiday City South lies entirely within Berkeley Township, according to the 2004se files. See below. Holiday City South is shown transparently overlayed on Berkely and Dover Townships, which are colored differently. -- JimIrwin 13:17, 23 December 2005 (UTC)

I'll accept that as proof. Back to Ocean County CDPs.... Alansohn 13:35, 23 December 2005 (UTC)

[edit] Matching CDP and Township names

There are many examples of CDPs and Townships with the same name. Most of these are what we've been calling "Trivial CDPs," where the CDP and Township are one and the same. In these cases, I have been ignoring the Township's CDP status. There are also a number of cases where the pair has the same name, but the two are not identical. I've seen this for Lakewood Township, New Jersey, and also for a bunch of pairs in Salem County: Alloway Township, New Jersey, Carneys Point Township, New Jersey and Pennsville Township, New Jersey. In all four of these cases, there had been two similarly named pages Lakewood Township, New Jersey (the township) and Lakewood, New Jersey (the CDP). In each case, I created a unique page for the CDP (e.g., Lakewood CDP, New Jersey) and changed Lakewood, New Jersey to a redirect to Lakewood Township, New Jersey. Many people had assumed that Lakewood, New Jersey was the page for the township and had been actively updating that page. All updates on the CDP pages were moved to the Township page, and all references to the Township being a CDP were removed from the Township pages. I will do the same if I find any other such pairs, and change any references on the County pages to link to the CDP pages. Alansohn 12:37, 27 December 2005 (UTC)

That's good. My opinion is that whenever the CDP and township are different, they should have their own pages, since they each have distinct statistics. I think I found one case last week when the CDP had been redirected to the township, and I reverted the CDP back to its own page. -- JimIrwin 16:11, 27 December 2005 (UTC)

[edit] Index map captions

The county index maps distinguish between a "Township" and an "Incorporated Place." Aren't we perpetuating the Census Bureau's miscategorization of New Jersey Townships as unincorporated places? Is there any way to mark the Incorporated Places as Boroughs / Cities / Towns / Villages, perhaps using a different color for each category? Alansohn 12:37, 27 December 2005 (UTC)

Actually, the census bureau includes townships as incorporated places (boroughs, townships, etc. are all "minor civil divisions"). The CDPs are the only unincorporated places in New Jersey tracked by the census bureau, outside of military reservations. Unfortunately, the dataset doesn't give me enough information to color all the various political entity types. It does allow me to distinguish townships from the rest, and since townships are generally larger, it was natural to color them differently. Also, the townships seem to be the older entities, and often totally surround embedded boroughs, villages, etc., especially in the more rural counties. It helps to highlight the boroughs by using a different color. I could change the legend for the non-townships to "other incorporated places".

It would have been nice to distinguish all five types of municipalities, but using "Other Incorporated Places" would be far better than "Incorporated Place" in dispelling the myth that Townships are unincorporated. Are you OK with the CDP changes? Alansohn 14:39, 27 December 2005 (UTC)

[edit] Highland Lake vs. Highland Lakes

I'm up to Sussex County (only two more to go regarding linking CDPs to their townships). There was a page for Highland Lake, New Jersey that was set up to redirect to Highland Lakes, New Jersey. I checked my data, your info and the Census Bureau web site, and they all had Highland Lake. So I switched the pages so that Highland Lakes, New Jersey is a redirect to the Highland Lake, New Jersey CDP page. It turns out that the USPS has a Highland Lakes post office in Sussex, which I assume overlaps with the Highland Lake CDP. These situations do crop up. I'm thinking of changing the Highland Lakes, New Jersey page to say it's the ZIP Code for 07422 in Sussex and have wikilinks to Vernon Township, New Jersey and reference the Highland Lake, New Jersey CDP page. Sound reasonable? Alansohn 14:39, 27 December 2005 (UTC)

"Highland Lakes" is more than a zip code, it is the name of a populated place (41°10'38"N 74°28'08"W) in the Geographic Names Information System (GNIS) of the USGS. "Highland Lake" is the name of the CDP which includes the region around Highland Lake (the lake). Normally the Census Bureau CDP name and the GNIS name correspond, but in this case they are different, probably because the CDP refers to the region around the lake. The GNIS database is the official place name database, and that is probably also why the USPS post office is "Highland Lakes". The USPS has its own method for choosing postal zones, and they won't correspond with any Census delineation. If there is a separate Wikipedia page for "Highland Lakes", it shouldn't have any census statistics or government information of its own, but it might have history and geographic information of its own. -- JimIrwin 16:11, 27 December 2005 (UTC)

[edit] Maps

Image:Hudson County, New Jersey Municipalities.png is a nice map, only if You could add some information like where the map data came from, and what tools you used to put it together -- for reference to others there (and in future maps too?). Thanks. feydey 17:37, 27 December 2005 (UTC)

OK. See the Lineage of index maps on my user page. -- JimIrwin 20:22, 27 December 2005 (UTC)

[edit] ZIP Codes and ZCTAs

Now that we have made it through MCDs and CDPs, there seems to be one more primary source of other unincorporated communities: ZIP Codes. While it seems that it's tough to get the ZIP Code mapping data from the USPS, perhaps we can use the Census Bureau's ZIP Code Tabulation Areas (ZCTAs), which are intended to correspond to the areas covered by the ZIP Codes. The relevant page from the Census Bureau seems to indicate that they have the TIGER/Line® files with updated national ZCTAs reflecting U.S. Postal Service ZIP Code changes through October, 2002. If this is the case, couldn't you do your magic to create maps showing the ZCTA and the relationship between the Zip Code / ZCTA and the parent municipality (or municipalities). Many ZIP Codes / ZCTAs will be trivial, one-to-one matches to municipalities. But there are loads of non-trivial relationships. Many people say that they "live" in their ZIP Code, not in their municipality. There are hundreds more Zip Codes than MCDs and CDPs combined. Any thoughts? Maybe trying a few samples? Alansohn 03:25, 30 December 2005 (UTC)

I'll give it a try, but my past experience with the zip code data in the TIGER/Line files is that it is incomplete, and generally omits rural areas. The main difficulty with zip codes are (1) the USPS doesn't create zip codes as geographic areas, and (2) they aren't fixed. Zip codes tend to match conventient delivery routes and can be quite convoluted. The USPS doesn't publish the information, and the people who have managed to create zip code outline datasets sell the information for profit. There are also constant changes to zip codes; perhaps several hundred each year get created or destroyed. There is a new 2005 first edition TIGER/Line that I just obtained, and I'll see what the ZCTAs look like in it. -- JimIrwin 12:42, 30 December 2005 (UTC)

I can only ask you to try. I already see that one of the ZIP Codes I was looking for (07710 for Adelphia) doesn't exist (probably as a rural area). But I do see that many of the places that've been on the most wanted list do exist on the file. If we include a caveat that any map represents the ZCTA and not the actual ZIP Code (something like: Please note that the ZIP Code area may not match the ZCTAs; ZCTAs were determined as of the 2000 Census and may be changed over time by the USPS; Please consult the USPS for any questions regarding the coverage area of a particular ZIP Code) we should at least be able to create pages for a majority of the "missing" areas in New Jersey. Furthermore, there is census data by ZCTA, so we should be able to create pages that look just like the CDP pages (ZIP Code Name is the area covered by ZIP Code 12345. The ZCTA for the area covers the communities of X, Y and Z in County Name. As of the United States 2000 Census, the ZCTA had a population of 1,234...). Let me know if you can tame this TIGER. Thanks again for all your work on the MCD and CDP maps! Aren't you due for another challenge. Alansohn 13:43, 30 December 2005 (UTC)

I analyzed the ZCTAs in and around Morris County (yellow tint), and am not pleased with the results. One, there are a lot of ZCTAs and I think it's going to be too much work for me to attempt to map each one. Second, zip codes don't follow political boundaries. In the map, I created light shading on the ZCTAs that span county boundaries. Each county is tinted separately in the map, with the lighter areas being the ZCTAs that span counties. Finally, the USPS assigns a primary name to a zip code, but sometimes accepts alternate names, and sometimes will not accept a common name. For example, ZCTA 07438 is called "Oak Ridge" by the Post Office, but it contains several communities, and more importantly, it includes much of Jefferson Township. Many if not most of the residents of Jefferson Township think of themselves as living in the township, not as living in Oak Ridge. See the image below.

Morris County and surrounding ZCTAs

[edit] Question about your maps for counties in New Jersey

I have a question about your map for Essex County that lists basically two categories: Incorporated Place and Township. This can be really confusing...

I guess I never fully understood this myself, but it was my understanding that in New Jersey all townships are considered "Incorporated" as per the Home Rule Act of 1917 (seeTownship (New Jersey)). Actually, this act places Cities, Villages, Towns, Townships, and Boroughs all on equal legal footing.

Effectively, it seems that there is no 'unincorportaed' place...

Roodog2k

You are correct, all the legal county subdivisions in New Jersey are incorporated. The main reason I distinquished between them is because the townships tend to be the larger entities, and the other incorporated places are often "islands" within the townships, and the color difference helps distinquish between them. In most counties, the other incorporated places are boroughs or towns and are rather small compared to the townships. I think on many of the county maps I used "Cities, towns, boroughs, villages" in the legend, as in Sussex County. I haven't gotten around to redoing the first couple of maps that were labeled like Essex County.

Thanks for reminding me to go back and clean them up.

JimIrwin 23:04, 28 January 2006 (UTC)

[edit] Oh, nice job!

Sorry... didn't mean the last message to sound critical...

Roodog2k [[User_talk:Roodog2k | (Hello there!)]] 17:33, 28 January 2006 (UTC)

[edit] Redundant Images

Hey, I noticed your redundant images were orphaned, so I listed them as candidates for speedy deletion. I originaly placed a warning template on your talk page for two of them before realizing you created the originals. --ThrashedParanoid 01:26, 1 February 2006 (UTC)

[edit] Really?

Did you really create that Maine map all by yourself?Gator (talk) 19:15, 17 February 2006 (UTC)

Yes, see Lineage of population maps for a description of the process. JimIrwin 21:29, 17 February 2006 (UTC)

[edit] Mislabled highway on map of Hawthorne, NJ

On the map you provided for Hawthorne, NJ, you labeled route 208 as route 20. Route 20 does not, at any point, traverse Hawthorne. (Route 20 ends just south of Hawthorne in Paterson.) If you would like to verify that, you can go to MapQuest and look up a map of Hawthorne and the surrounding area.

I noticed the error only because I used to live in Hawthorne and traveled both route 208 and route 20 on almost a daily basis.

• lithium6ion 23:07, 18 Feb 2006 (UTC)

It has been corrected. Thanks! --JimIrwin 11:40, 19 February 2006 (UTC)

[edit] Hello Jim

Great maps. I am working on the county templates, and my personal interest is in the unincorporated historical enclaves. How shoud I go about finding the borders of places like Six Mile Run, New Jersey and putting it on a map? I see someone has asked a similar question. --Richard Arthur Norton (1958- ) 01:27, 27 February 2006 (UTC)

I'm not sure where you would get the information. It could be that places like Six Mile Run never had formal boundaries. The Census Bureau data that I used only has boundaries of legal entities and census-designated-places. I would definitely not use zip code boundaries, since zip code locations only reflect the logistics of the Postal Service and certainly have no historical basis.

You can find a point location for Six Mile Run from the Geographic Names Information Service (GNIS). The GNIS provides a single geographic point for mapping purposes. The GNIS is the official source of geographic names in the US, and seems to say that Six Mile Run is really called "Sixmile Run." The point location would allow you to create a dot-on-a-map locator.

-- JimIrwin 13:49, 27 February 2006 (UTC)

[edit] Railway map?

I'm wondering if you'd be interested in creating a map for the New York, Susquehanna and Western Railway article. The article is sort of confusing for someone who doesn't know the route through NJ, or elsewhere for that matter. Would you be able to make an entire route map, a route map in sections, or just a NJ map? Thanks for considering!
Kether83 05:04, 8 October 2006 (UTC)

Unfortunately I don't have any spare time at the moment, otherwise it would be interesting. The current route is likely available in the Bureau of Transporation Statistics NTAD dataset. I wouldn't know where to find the data for historical routes.

-- JimIrwin 11:59, 11 October 2006 (UTC)

[edit] Maps Query

what program do you use to create the maps? i.e. like the one at the Stillwater Township, New Jersey article? —ExplorerCDT 17:11, 27 October 2006 (UTC)

Those were done with Manifold using several public datasets. See the lineage of place maps description. —JimIrwin 11:03, 28 October 2006 (UTC)

[edit] Paulins Kill

Jim, Thanks for your prompt answer to my above question. Could you put together a map of the Paulins Kill, a tributary of the Delaware River, in Northwestern New Jersey for the article? If you need any information aside from that in the article, I'd be happy to provide it. I am aiming to nominate the article for featured article status, and several people have mentioned that a map is needed on the article. If you have the time, I'd be much obliged.—ExplorerCDT 20:17, 28 October 2006 (UTC)

I'll take a look at it. It may be a week or so. —JimIrwin 12:13, 31 October 2006 (UTC)

Take your time, no rush. Right now the article is under peer review, so we have a little while before I recommend it for FA status. Thanks. —ExplorerCDT 15:13, 31 October 2006 (UTC)

User:Kmusser created a map image and put on the article only moments after I posted the above. So, that is taken care of and you're off the hook. Thanks again though. —ExplorerCDT 17:02, 31 October 2006 (UTC)

Jim, I just nominated your map for the Paulins Kill for Featured Pictures. The nomination can be found here: Wikipedia:Featured_picture_candidates/Map_of_the_Paulins_Kill_and_its_Watershed. Simultaneously, Paulins Kill is a candidate for Featured Article status.

Again, I thank you. Your work has helped to make a great article. —ExplorerCDT 00:48, 22 December 2006 (UTC)

Thanks. I probably won't have the time to address the flood of comments that are sure to arrive, at least in the near term. People looking at it through the eyes of professional cartographers will recognize it for amateur effort that it is. I appreciate your thanks. Perhaps others will be motivated to improve it though, and that will make it worth the nomination. — JimIrwin 12:59, 22 December 2006 (UTC)

[edit] New SVG maps

Jim, I just noticed that you've uploaded some new SVG maps! They look great, thank you. Have you changed the software you use to render the maps? Can older PNG maps be replaced with SVG versions easily? Thanks. --ChrisRuvolo (t) 20:55, 28 December 2006 (UTC)

I've acquired a program that takes the output of my GIS software and converts it to SVG. Unfortunately, the SVG requires some manual editing before it can be displayed, so it is not quite "easy" to convert the older PNG files to SVG versions. The two maps that I added today were essentially experiments. It takes perhaps another 10-20 minutes of work to generate SVG for a new map, but compared to the amount of work it takes to create the new map itself, it is relatively little additional effort. However, going back and reworking prior maps would consume quite a bit of time. I intend to generate SVG for any new maps I create, as long as the SVG is reasonably small. — JimIrwin 23:01, 28 December 2006 (UTC)

Great maps on the Geography of New Jersey article. Thanks again! —ExplorerCDT 01:13, 30 December 2006 (UTC)

[edit] Paulins Kill now a Featured Article

As of a few moments ago, Paulins Kill was promoted to Featured Article status. I just wanted to thank you for your contributions to and suggestions for improving the article over these past few months and that I appreciate your help in bringing this article to notice as a Featured Article. Once again, thank you, and keep up the good work. —ExplorerCDT 22:46, 10 January 2007 (UTC)

[edit] Population maps

You wrote in (c) that you used custom Perl Script to calculate the population per square mile. Would it be possible to obtain this script? SanitySolipsism 15:58, 9 February 2007 (UTC)

More generally, could you be a bit more specific as to what data you used on the map (the urls for the data and what fields correspond to what data) SanitySolipsism 19:10, 9 February 2007 (UTC)

[edit] New Hampshire's population density map has an error

The map legend and map don't agree. The high density areas of New Hampshire are in the S.E. corner, while the low density areas are in the north. This map has it completely backwards. From the looks of other maps it appears though the legend is right, but the map coloration is wrong. See: Link to map here (talk) 12:57, 9 March 2007 (UTC)

There is no error. The map shows the high density in the S.E. corner, just as you say it should. The redder colors are the higher density areas, the greener colors are the lower density areas. JimIrwin 14:53, 9 March 2007 (UTC)

Oops! I'm sorry I am colorblind. I was just judging by darker to lighter shades and it looked like the darker shade in the north corresponded to the bottom of the density chart. Now that I look closer I can see where I was mistaken.

FYI- Red and green are particularly poor colors to choose in map-making or charts, especially on the web. Red/Green Color blindness affects up to 10% of all males in the U.S. and thus it makes it difficult for these people to discern subtleties that are required when maps and charts use these colors. Single color variation from light to dark, or dual coloration from yellow to blue would be much better. There are a number of resources on the internet to help people make images, maps, and charts more friendly to colorblind people. You seem like a person interested in geography and map making, so this subject might be something for you to consider in the future. (One of my best friends has a masters in geography, and I love maps and geography too!). Take care. Vivaldi (talk) 09:08, 16 March 2007 (UTC)

I figured as much. Unfortunately, colors that work well for color blind people can often be astonishingly ugly to people who are not color blind. Blue-yellow is one of those combinations, in my opinion.