topic Re: Reprojecting to Web Mercator in Mapping Questions

Reprojecting to Web Mercator

DavidRobinson7 — Wed, 24 Aug 2016 20:26:25 GMT

I have the Census TIGER/Line state boundaries, which I am trying to reproject to resemble more commonly held perceptions of state boundaries (i.e. Google maps). The TIGER/Line data is currently projected in:

Projected Coordinate System: Hawaii_Albers_Equal_Area_Conic
Projection: Albers

I am trying to reproject to WGS 1984 Web Mercator (auxiliary sphere), which from my research I understand to be the projection used by Google.

The reprojection tool runs successfully; however, the output looks the same as the input (stretched boundaries as compared to Google maps). I've spent an hour or two researching the best way to accomplish this particular task, but I'm not coming up with answers that have worked.

Advice is much appreciated!

Re: Reprojecting to Web Mercator

DanPatterson_Retired — Wed, 24 Aug 2016 20:32:14 GMT

well it will look the same if you keep everything in the same dataframe. The data frame uses the coordinate system of the first layer added, then reprojects everything else added on the fly to match... always add projected data to a new dataframe to see what it looks like, or set the current data frame to what you want to view it in.

Re: Reprojecting to Web Mercator

DavidRobinson7 — Wed, 24 Aug 2016 20:40:29 GMT

Thanks... but I have already tried adding the output to a new application (new dataframe), with no changes.

(sorry about the double post, new to this community)

Re: Reprojecting to Web Mercator

MelitaKennedy — Wed, 24 Aug 2016 23:20:45 GMT

Just double-checking that you used the Project tool and not the Define Projection Tool.

How are you checking it against Google Maps?

Re: Reprojecting to Web Mercator

DavidRobinson7 — Thu, 25 Aug 2016 11:17:06 GMT

Thanks.. Yes I used the Project tool.

Visually, the output is the same as the input (Equal Area Conic projection) and not Mercator.

Re: Reprojecting to Web Mercator

DanPatterson_Retired — Thu, 25 Aug 2016 11:44:30 GMT

in separate data frames? not in the same data frame? as per my original post, because of projection-on-the-fly

Re: Reprojecting to Web Mercator

DavidRobinson7 — Thu, 25 Aug 2016 11:52:38 GMT

No, I visually compared the input and output in separate data frames. I added the output to a new data frame so on-the-fly projection would not occur.

Re: Reprojecting to Web Mercator

DanPatterson_Retired — Thu, 25 Aug 2016 12:00:41 GMT

this then goes back to the original data, which is in Albers'... what are the extents? and was it Projected to Albers from a geographic coordinate system (aka decimal degrees) or did it come that way? I would be surprised if a gov't department would ship already projected data rather than data in a GCS

Re: Reprojecting to Web Mercator

DavidRobinson7 — Thu, 25 Aug 2016 12:04:43 GMT

Thanks for the help. Listed below are the details related to the original data:

Projected Coordinate System: Hawaii_Albers_Equal_Area_Conic
Projection: Albers

False_Easting: 0.00000000
False_Northing: 0.00000000
Central_Meridian: -157.00000000
Standard_Parallel_1: 8.00000000
Standard_Parallel_2: 18.00000000
Latitude_Of_Origin: 13.00000000
Linear Unit: Meter

Geographic Coordinate System: GCS_North_American_1983
Datum: D_North_American_1983
Prime Meridian: Greenwich
Angular Unit: Degree

Re: Reprojecting to Web Mercator

DanPatterson_Retired — Thu, 25 Aug 2016 12:12:24 GMT

MKennedy-esristaff‌ ... come on back

If it doesn't look right, in Albers and the coordinate system is right, then we shall await for a response as to the root of the problem Stretched to me means that the coordinates are geographic and there is no projection applied, but I reside at much higher latitudes where we use lambert CC instead of alber.

Re: Reprojecting to Web Mercator

DavidRobinson7 — Thu, 25 Aug 2016 12:25:16 GMT

Thanks. As I mentioned, the reprojection completes successfully. But the shape of the features in the output is the same as the input (even when adding to a new data frame). The details of the input are listed below, if helpful:

Projected Coordinate System: WGS_1984_Web_Mercator_Auxiliary_Sphere
Projection: Mercator_Auxiliary_Sphere
False_Easting: 0.00000000
False_Northing: 0.00000000
Central_Meridian: 0.00000000
Standard_Parallel_1: 0.00000000
Auxiliary_Sphere_Type: 0.00000000
Linear Unit: Meter

Geographic Coordinate System: GCS_WGS_1984
Datum: D_WGS_1984
Prime Meridian: Greenwich
Angular Unit: Degree

I studied projections and coordinates systems in great detail when I was in graduate school, but alas that was many moons ago and I hate to admit I didn't retain all of the knowledge - so I'm struggling to figure out where/ what the issue is.

Thanks for the help.

Re: Reprojecting to Web Mercator

NeilAyres — Thu, 25 Aug 2016 13:23:06 GMT

Although it has little relevance to this debate, I thought I would have a look at the 2 projections. Centred on Hawaii, I think that this is where we are talking about.

Both frames are exactly the same scale and are centred on the same point. Note that Web Mercator the scale is smaller. It really is a bad projection to use for almost anywhere!!!!

Re: Reprojecting to Web Mercator

DavidRobinson7 — Thu, 25 Aug 2016 13:54:55 GMT

Thanks. I don't know why the original data is projected into Hawaii_Albers_Equal_Area_Conic, as the dataset covers the contiguous US (probably some reason why). But, I'm trying to create a US state boundaries dataset that resembles Google maps.

Re: Reprojecting to Web Mercator

NeilAyres — Thu, 25 Aug 2016 16:23:05 GMT

Well a dataset covering the entire USA would look a little funny using the Hawaii projection (centred on 157degW). Isn't the whole of eastern USA rather distorted?

Re: Reprojecting to Web Mercator

DavidRobinson7 — Thu, 25 Aug 2016 17:01:58 GMT

I do not know why the original data (as provided by Census TIGER/Line) is projected in the Hawaii Albers projection - that is the original projection - my issue is trying to reproject it to Web Mercator.

Re: Reprojecting to Web Mercator

DanPatterson_Retired — Thu, 25 Aug 2016 17:47:36 GMT

have you checked to see whether you can download the data in another format? You have two non-nationals trying to help with US census data which I am sure someone could step in and solve the problem in a snap. Trying to reproject something that is suitable to something that is barely tolerable is an onerous task.

Re: Reprojecting to Web Mercator

DavidRobinson7 — Thu, 25 Aug 2016 17:59:15 GMT

I have not been able to find the data I am looking for in the projection I've described. While one may deem it to be an onerous task, 1) I was assigned to find the state boundaries projected as described, and 2) Being that my multiple attempts to reproject the dataset without the desired results, I am now interested in learning why the desired result is not being produced and how I may resolve the issue.

Thanks.

Re: Reprojecting to Web Mercator

DanPatterson_Retired — Thu, 25 Aug 2016 18:08:02 GMT

Your life would be easier if you could get the data in decimal degrees (aka a GCS) and project it directly to the web mercator... that is why I ask.

Re: Reprojecting to Web Mercator

DavidRobinson7 — Thu, 25 Aug 2016 18:09:39 GMT

Ok, I will try to find the data in a geographic coordinate system. Thanks for the tip.

Re: Reprojecting to Web Mercator

MelitaKennedy — Fri, 26 Aug 2016 18:57:38 GMT

Hi David,

Is the location you downloaded the data from public? That is, can you post the location so someone else can try downloading the data?

Also, would you mind sharing the extent of the original data? If you add the data to a new ArcMap and don't change the data frame's coordinate system, what are the extents as you check the edges of the data?

If I add the Esri topographic base map (EPSG:3857), and change the data frame's coordinate system to Hawaii Albers, the extents of the lower 48 are (very roughly!):

x min: 2950000 m

y min: 2060000

x max: 8400000

y max: 5000000

(As far as I know, US census data is almost always in geographic, latitude-longitude).