Changing a grid's scale_factor

[PaulWessel]

Description of the problem

Not sure if this is a bug or not but it is certainly not straightforward to do or figure out, and I should have an inside track on this...

I have a netCDF short int grid with values that need to be scaled by 0.01 to be used. OK, so I can plot it as grid.nc+s0.01. But this data will never be used as is, always with that scale, so better to insert the 0.01 scale into the grid. I have tried various schemes with grdedit and grdconvert but no success. Making this issue mostly so it is not forgotten and once a solution (bug fix, improved documentation) happens then we can delete it.

One can experiment with this grid whose z values should be 0-100 on output, but that requires +s0.01 when used.

gmt grdmath -R0/10/0/10 -I1 X Y MUL 100 MUL = t.nc=ns

[seisman]

Does gmt grdedit -D+s0.01 input.grd work?

[seisman]

Just find a potential bug with grdedit. Follow the steps below to reproduce it:

1. Download a cache data for testing

$ gmt which -Gl @AFR.nc

2. Check grid information

$ gmt grdinfo AFR.nc
AFR.nc: Title: Produced by grdcut
AFR.nc: Command: grdcut etopo2m.nc -R0/10/0/10 -GAFR.nc
AFR.nc: Remark:
AFR.nc: Pixel node registration used [Geographic grid]
AFR.nc: Grid file format: nf = GMT netCDF format (32-bit float), CF-1.7
AFR.nc: x_min: 0 x_max: 10 x_inc: 0.0333333333333 (2 min) name: longitude n_columns: 300
AFR.nc: y_min: 3.46944695195e-18 y_max: 10 y_inc: 0.0333333333333 (2 min) name: latitude n_rows: 300
AFR.nc: z_min: -5023 z_max: 3614 name: z
AFR.nc: scale_factor: 1 add_offset: 0
AFR.nc: format: netCDF-4 chunk_size: 150,150 shuffle: on deflation_level: 3

3. Run grdedit and save it to a new grid

$ gmt grdedit AFR.nc -D+s10 -GAFR10.nc

4. Run grdedit but overwrite the input grid

$ gmt grdedit AFR.nc -D+s10

5. Check grid information of AFR.nc and AFR10.nc

$ gmt grdinfo AFR.nc
AFR.nc: Title: Produced by grdcut
AFR.nc: Command: grdcut etopo2m.nc -R0/10/0/10 -GAFR.nc
AFR.nc: Remark:
AFR.nc: Pixel node registration used [Geographic grid]
AFR.nc: Grid file format: nf = GMT netCDF format (32-bit float), CF-1.7
AFR.nc: x_min: 0 x_max: 10 x_inc: 0.0333333333333 (2 min) name: longitude n_columns: 300
AFR.nc: y_min: 6.93889390391e-18 y_max: 10 y_inc: 0.0333333333333 (2 min) name: latitude n_rows: 300
AFR.nc: z_min: -502300 z_max: 361400 name: z
AFR.nc: scale_factor: 10 add_offset: 0 packed z-range: [-50230,36140]
AFR.nc: format: netCDF-4 chunk_size: 150,150 shuffle: on deflation_level: 3

scale_factor is 10 in the grid, but the z_min and z_max is 100 times of the original data. Why?

$ gmt grdinfo AFR10.nc
AFR10.nc: Title: Produced by grdcut
AFR10.nc: Command: grdcut etopo2m.nc -R0/10/0/10 -GAFR.nc
AFR10.nc: Remark:
AFR10.nc: Pixel node registration used [Geographic grid]
AFR10.nc: Grid file format: nf = GMT netCDF format (32-bit float), CF-1.7
AFR10.nc: x_min: 0 x_max: 10 x_inc: 0.0333333333333 (2 min) name: longitude n_columns: 300
AFR10.nc: y_min: 6.93889390391e-18 y_max: 10 y_inc: 0.0333333333333 (2 min) name: latitude n_rows: 300
AFR10.nc: z_min: -5023 z_max: 3614 name: z
AFR10.nc: scale_factor: 1 add_offset: 0
AFR10.nc: format: netCDF-4 chunk_size: 150,150 shuffle: on deflation_level: 3

scale_factor is still 1.

[PaulWessel]

In my case the grdedit -D+s0.01 works exactly as I want, but if I want to write that to a new grid via -Gnew.grd the new grid is not correct.
Your AFR case behaves the same way. Yes, the xin/max is affected since you are scaling the data by 10. In my case I want to scale by 1/100 and the zmax changes accordingly, which is what I want. So in your case, when you use your grid you recover correct depths since your data is now stored in decimeters...

[seisman]

So in your case, when you use your grid you recover correct depths since your data is now stored in decimeters...

AFR.nc: z_min: -502300 z_max: 361400 name: z
AFR.nc: scale_factor: 10 add_offset: 0 packed z-range: [-50230,36140]

If I understand it correctly, the z_min and z_max should be -50230 and 36140, not -502300 and 361400. If I use grd2xyz to output the data, I should see the z value in the range -5023 to 3614, right?

[PaulWessel]

Yes, there is clearly some problems here:

gmt which -Gl @AFR.nc
echo 5 5 | gmt grdtrack -GAFR.nc -nn
5	5	-79
gmt grdedit AFR.nc -D+s10
echo 5 5 | gmt grdtrack -GAFR.nc -nn
5	5	-790

We getting the factor of 10 twice. However, my other case seems to work:

gmt grdmath -R0/10/0/10 -I1 X Y MUL 100 MUL = t.nc=ns
echo 10 10 | gmt grdtrack -Gt.nc -nn
10	10	10000
gmt grdedit t.nc -D+s0.01
echo 10 10 | gmt grdtrack -Gt.nc -nn
10	10	100

[stale]

This issue has been automatically marked as stale because it has not had activity in the last 90 days. It will be closed if no further activity occurs within 7 days. Thank you for your contributions.