topic Re: project MODIS Burned Area Product in Data Management Questions

project MODIS Burned Area Product

CarieErnst1 — Wed, 17 Nov 2010 14:54:13 GMT

Hello,

I can't seem to get anywhere on this and am hoping that Melita might be able to respond since it seems she has been dealing with this a lot, but any help is appreciated.

I have GEOTIFF's provided by the University of Maryland for MODIS burned area for years 2003-2009. I am trying to take it from a Platte Carree projection to Albers Equal Area. please help me!

thanks to all, Carie

Re: project MODIS Burned Area Product

MelitaKennedy — Wed, 17 Nov 2010 21:50:52 GMT

Hello,

I can't seem to get anywhere on this and am hoping that Melita might be able to respond since it seems she has been dealing with this a lot, but any help is appreciated.

I have GEOTIFF's provided by the University of Maryland for MODIS burned area for years 2003-2009. I am trying to take it from a Platte Carree projection to Albers Equal Area. please help me!

thanks to all, Carie

Hi Carie,

What problem are you encountering? Does the original, downloaded GeoTIFF have a defined coordinate system? Does the data appear to be georeferenced (the raster extents aren't equal to 'cell' counts)?

If georeferenced but doesn't have a defined coordinate system, what happens when you define its coordinate system? The online documentation is less than helpful about the coordinate system details, isn't it?

I'm assuming that the GCS/Datum is the same as listed in the report. Because I can't attached .prj files, here's the contents of the a .prj file:

GEOGCS["GCS_MODIS_Burned_Area_Sphere",DATUM["D_MODIS_Burned_Area_Sphere",SPHEROID["MODIS_Burned_Area_Sphere",6371007.181,0.0]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]]

Place it in a text file (no word wrapping, single line) with a .prj extension and place it in coordinate systems, geographic coordinate systems. Now you can create a custom projected coordinate system with this as its geographic coordinate system. Just leave the Plate Carree parameters as-is. I assume the central meridian is zero.

Melita

Re: project MODIS Burned Area Product

CarieErnst1 — Thu, 18 Nov 2010 14:03:56 GMT

Melita,
    Thank you for your response. Yes, you are correct that the original GEOTIFF is georeferenced, but comes without a spatial reference. Based on the Burned Area Manual, I know that it has been reprojected from Sinusoidal into Platte Carree and I have no issue defining the GEOTIFF as such. But, I need a projection that preserves accurate area, thus Albers Equal Area for the CONUS is ideal. this is what I've tried:

1) in ArcCatalog or in ArcMap (I've tried both) I go to the properties of the file and give it the Platte Carre projection (WGS 1984 by default) I've also now tried assigning it your MODIS sphere proj.

2) Next, I go to arctools > data management > projections and transformations > project raster. For the input I select my file, it shows up with the projection I just defined. For the Output I select projected coord system > continental > North America > USA contiguous Albers Equal Area Conic.prj. Then the only thing I modify is the latitude of origin, changing it to 0.here it is:

Projection: Albers
False_Easting: 0.000000
False_Northing: 0.000000
Central_Meridian: -96.000000
Standard_Parallel_1: 29.500000
Standard_Parallel_2: 45.500000
Latitude_Of_Origin: 0.000000
Linear Unit: Meter (1.000000)

Geographic Coordinate System: GCS_North_American_1983
Angular Unit: Degree (0.017453292519943299)
Prime Meridian: Greenwich (0.000000000000000000)
Datum: D_North_American_1983
Spheroid: GRS_1980
    Semimajor Axis: 6378137.000000000000000000
    Semiminor Axis: 6356752.314140356100000000
    Inverse Flattening: 298.257222101000020000

3) as soon as I click ok, the program shuts down, every time.

It seems the issue is related to the difference in spheroids between WGS 1984 or MODIS sphere and the NAD_1983 sphere. And I know there are some transformations for them, but I can't get them to work.

The other part of my problem is that I'm trying to get this burned area to overlay with a cropland data layer I have for the conUS. here is the projection specs for my cropland data layer. This is not the original projection, it was reprojected to this via ArcInfo workstation, which as I understand assigns the clarke 1866 as a default. I was given this layer this way and have to make the two overlay accurately to determine how much burned area is cropland.

Projection: Albers
False_Easting: 0.000000
False_Northing: 0.000000
Central_Meridian: -96.000000
Standard_Parallel_1: 29.500000
Standard_Parallel_2: 49.500000
Latitude_Of_Origin: 0.000000
Linear Unit: Meter (1.000000)

Geographic Coordinate System: GCS_Clarke_1866
Angular Unit: Degree (0.017453292519943299)
Prime Meridian: Greenwich (0.000000000000000000)
Datum: D_Clarke_1866
Spheroid: Clarke_1866
    Semimajor Axis: 6378206.400000000400000000
    Semiminor Axis: 6356583.799998980900000000
    Inverse Flattening: 294.978698200000000000

So, I've also been trying to get the Burned Area geotiffs into a Clarke_1866 albers, but I have no experience with ArcInfo Workstation and haven't been able to reproduce what my boss did with the cropland data layer.

This must seem like such a mess, at least it does to me, hopefully not to you! The bottom line is that I need to get those GEOTIFF's into an alber's projection that will overlay with the cropland data. Is this possible? Can you think of any way to do so? Again, thank you, Melita.

Carie

Re: project MODIS Burned Area Product

MelitaKennedy — Thu, 18 Nov 2010 19:24:29 GMT

Melita,
    Thank you for your response. Yes, you are correct that the original GEOTIFF is georeferenced, but comes without a spatial reference. Based on the Burned Area Manual, I know that it has been reprojected from Sinusoidal into Platte Carree and I have no issue defining the GEOTIFF as such. But, I need a projection that preserves accurate area, thus Albers Equal Area for the CONUS is ideal. this is what I've tried:

1) in ArcCatalog or in ArcMap (I've tried both) I go to the properties of the file and give it the Platte Carre projection (WGS 1984 by default) I've also now tried assigning it your MODIS sphere proj.

2) Next, I go to arctools > data management > projections and transformations > project raster. For the input I select my file, it shows up with the projection I just defined. For the Output I select projected coord system > continental > North America > USA contiguous Albers Equal Area Conic.prj. Then the only thing I modify is the latitude of origin, changing it to 0.here it is:

Projection: Albers
False_Easting: 0.000000
False_Northing: 0.000000
Central_Meridian: -96.000000
Standard_Parallel_1: 29.500000
Standard_Parallel_2: 45.500000
Latitude_Of_Origin: 0.000000
Linear Unit: Meter (1.000000)

Geographic Coordinate System: GCS_North_American_1983
Angular Unit: Degree (0.017453292519943299)
Prime Meridian: Greenwich (0.000000000000000000)
Datum: D_North_American_1983
Spheroid: GRS_1980
    Semimajor Axis: 6378137.000000000000000000
    Semiminor Axis: 6356752.314140356100000000
    Inverse Flattening: 298.257222101000020000

3) as soon as I click ok, the program shuts down, every time.

It seems the issue is related to the difference in spheroids between WGS 1984 or MODIS sphere and the NAD_1983 sphere. And I know there are some transformations for them, but I can't get them to work.

The other part of my problem is that I'm trying to get this burned area to overlay with a cropland data layer I have for the conUS. here is the projection specs for my cropland data layer. This is not the original projection, it was reprojected to this via ArcInfo workstation, which as I understand assigns the clarke 1866 as a default. I was given this layer this way and have to make the two overlay accurately to determine how much burned area is cropland.

Projection: Albers
False_Easting: 0.000000
False_Northing: 0.000000
Central_Meridian: -96.000000
Standard_Parallel_1: 29.500000
Standard_Parallel_2: 49.500000
Latitude_Of_Origin: 0.000000
Linear Unit: Meter (1.000000)

Geographic Coordinate System: GCS_Clarke_1866
Angular Unit: Degree (0.017453292519943299)
Prime Meridian: Greenwich (0.000000000000000000)
Datum: D_Clarke_1866
Spheroid: Clarke_1866
    Semimajor Axis: 6378206.400000000400000000
    Semiminor Axis: 6356583.799998980900000000
    Inverse Flattening: 294.978698200000000000

So, I've also been trying to get the Burned Area geotiffs into a Clarke_1866 albers, but I have no experience with ArcInfo Workstation and haven't been able to reproduce what my boss did with the cropland data layer.

This must seem like such a mess, at least it does to me, hopefully not to you! The bottom line is that I need to get those GEOTIFF's into an alber's projection that will overlay with the cropland data. Is this possible? Can you think of any way to do so? Again, thank you, Melita.

Carie

Hi Carie,

Oh yes, transformations. First, can you overlay the Burned Area data with your cropland data in ArcMap? Ignore any warning messages about different geographic coordinate systems/datums.

If it doesn't overlay, try defining a custom transformation.

Data frame properties, coordinate systems tab, Transformations, New button. In the New dialog, change the method to Geocentric Translation, leave the parameters set to zeroes. OK all and recheck the overlay.

Melita

Re: project MODIS Burned Area Product

CarieErnst1 — Tue, 23 Nov 2010 14:09:23 GMT

Melita,
They do overlay if I ignore the geographic coordinate system warning. The data frame is set to the MODIS sphere you provided me with, but it's still unprojected. I've tried projecting it to two different custom projections 1) GCS MODIS sphere, albers projection specs 2) GCS Clarke 1866, albers projection specs. (albers specs: central meridian -96.0, std parallel 1 29.5, std parallel 2 45.5, lat of origin 0). Now I just need to decide which one to go with.

With your help I think I am at least on the right track. I won't hesitate to contact you again if I get stuck (and probably will). Thank you for the help, Carie