Reprojecting to match USGS LAEA_Sphere_19 eMODIS/AVHRR data?

[ZacharyHart]

Completely stumped here. I have seen a handful of somewhat similar threads dealing with Lambert Azimuthal Equal Area...the problem here is working out the proper transformation. Here's a bit of a summary of what I do know.

• This (NDVI) data can be obtained from USGS here [AVHRR in this example].
• I have other data I am trying to align with this data (GCS_North_American_1983 WKID: 4269).
• There is no WKID associated with the data's projection info, nor is there a GCS or datum.
• When searching for 'LAEA_Sphere_19', surprisingly, only a handful of results are returned. One result provides some additional insight. Spatial Reference .org lists this as "SR-ORG Projection 8750 - eModis Phenology"
• The ESRI WKT reflects exactly what we see in the raster's spatial reference properties:

  PROJCS["LAEA_Sphere_19",GEOGCS["GCS_unknown",DATUM["D_unknown",SPHEROID["Sphere",6370997,0]],PRIMEM["Greenwich",0],UNIT["Degree",0.017453292519943295]],PROJECTION["Lambert_Azimuthal_Equal_Area"],PARAMETER["latitude_of_origin",45],PARAMETER["central_meridian",-100],PARAMETER["false_easting",0],PARAMETER["false_northing",0],UNIT["Meter",1]]

• The raster metadata provides no additional insight.
• There is much discussion (and stories of headache) in dealing with the Sinusoidal projection associated with other USGS/NASA MODIS data, but this is not that projection. Out of desperation, I tried creating the custom geographic transformation found here...of course that failed to yield proper results. This was another dead-end waste of time.
• The various online tools associated with manipulating the USGS/NASA HDF files are not valid for this data. This are TIFs for NDVI Phenology.
• My source data is GCS_North_American_1983; i need to somehow match to this LAEA Sphere 19.

Please, any insight or help here would be most appreciated...I cannot believe how little documentation there is regarding this!

[LukasFalk · Accepted solution]

Hey Zach, 

Unfortunately there are no geographic transformations between a sphere and a spheroid/datum.  You can find information on this at the URL below. 

https://support.esri.com/en/Technical-Article/000009008

The only workaround in this situation would be to redefine the Geographic Coordinate System for the data you want to project to the same sphere as the AVHRR data, then project it from Geographic to LAEA_Sphere_19.  The other alternative would be to georeference the data to the same coordinate system as the AVHRR data manually.  

This kind of thing comes up a lot.  The history behind it is ArcInfo Workstation only supported the LAEA projection on a sphere, not on a datum/spheroid.  The US government is still using this coordinate system for Modis and other weather related data.  

[ZacharyHart]

Thanks Lukas,

Now i need to wrap my head around the suggested workaround as I'm not quite clear on that yet...my brain is fried from spending days trying to research this problem.

[LukasFalk]

Hey Zach, 

You should be able to input the rasters you want projected into the Define Projection tool and reference the AVHRR data to get them into the same sphere.  Then use the project tool to assign the LAEA_Sphere_19 as their PCS.  

You could also look into the Geographic Coordinate Systems > Spheroid based as the GCS in the Define Projection tool.  The name Spheroid is a bit misleading because these are actually spheres.  Just make sure the semi-major axis is the same length as the LAEA_Sphere (6370997.0000).