Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

Inaccurate terrain height

J

jfqbsm

Member
Hello,
I created the initial conditions for a limited region in the Alps, with a variable mesh size of 60-3km. By checking the coordinates of the cells with the altitudes of the ground I notice that these are not correct.
I compared the value of the ter and zgrid fields of the init.nc file with the real value and I get significant differences. For example:
the cell in the center of the region I have 1463m instead of 1070m,
the maximum value of the ter field of all the cells is 3848m while I should approach 4800m.
What could explain these differences?
 
In principle, the value the 'ter' field in an MPAS-A grid cell should represent the average terrain height over the entire grid cell. So in cases where there is a mountain peak at the center of a cell, for example, we would expect that the value of the 'ter' field for that cell would be less than the height of the peak, and the reverse would be true in cases where there is a deep valley at the center of a grid cell.

The 'var2d' field gives the standard deviation of the sub-grid-scale terrain elevation, and it could be interesting to check the value of this field for grid cells whose average terrain elevation differs significantly from point measurements of actual terrain elevation.

Note that both the 'ter' and 'var2d' fields are computed from the GMTED2010 30-arc-second terrain elevation dataset.
 
mgduda said:
In principle, the value the 'ter' field in an MPAS-A grid cell should represent the average terrain height over the entire grid cell. So in cases where there is a mountain peak at the center of a cell, for example, we would expect that the value of the 'ter' field for that cell would be less than the height of the peak, and the reverse would be true in cases where there is a deep valley at the center of a grid cell.

The 'var2d' field gives the standard deviation of the sub-grid-scale terrain elevation, and it could be interesting to check the value of this field for grid cells whose average terrain elevation differs significantly from point measurements of actual terrain elevation.

Note that both the 'ter' and 'var2d' fields are computed from the GMTED2010 30-arc-second terrain elevation dataset.
Click to expand...
Thanks for your answer,
the var2d give me 316m for the first cell and 268m for the second.
So i have a second question: what could be the best way to interpolate for example the t2m variable?
 
haiqingsong said:
Try SRTM DEM.
Click to expand...
Hello, could you please take a look at the problem that I use SRTM elevation data about the mismatch between terrain height and actual height, and can you give me some suggestions for solving it? Please refer to my post for specific details, thank you very much！
 
Kang Huang said:
Hello, could you please take a look at the problem that I use SRTM elevation data about the mismatch between terrain height and actual height, and can you give me some suggestions for solving it? Please refer to my post for specific details, thank you very much！
Click to expand...
Where's your post?
 
Hello,
i'm not sure this is valid for MPAS ? There is no option to use another dataset than gmted2010_30s. I've an equivalent nested domain in WRF where i can more or less match the coordinate point of interest with the grid center and so the terrain height is close to reality, and my forecasting is not so bad.
With MPAS the cells coordinate are fixed.
 
You must log in or register to reply here.
Top