liu-jincan / study-fpm Goto Github PK

View Code? Open in Web Editor NEW

0.0 0.0 0.0 7.32 MB

学习fortran-package-manager(fpm)的过程记录~~

Fortran 95.26% C 0.19% CMake 4.55%

study-fpm's People

Contributors

Watchers

study-fpm's Issues

Global mesh

First of all, thank you so much for developing this tool, it is very useful!

Is your feature request related to a problem? Please describe.

I want to create a globalized unstructured grid for WAVEWATCH3 to run, and the code to generate the grid is as follows (using the Example_7 program as a template):

min_el    = 20e3;  	             % pdf 3.1, minimum resolution in meters.
bbox      = [-180 180;-89 90];   % pdf 2.4, lon min lon max; lat min lat max
max_el    = 120e3; 		         % pdf 3.1, maximum resolution in meters. 
wl        = 30;                  % pdf 3.4, 30 elements resolve M2 wavelength.
dt        = 0;                   % pdf 4.2, Only reduces res away from coast
grade     = 0.35;                % pdf 4.1, mesh grade in decimal percent. 
R         = 3; 			         % pdf 3.3.2, Number of elements to resolve feature.
slp       = 10;                  % pdf 3.5, slope of 10
dem       = 'SRTM15+.nc';        % 
coastline = {'GSHHS_f_L1','GSHHS_f_L6'};
gdat1 = geodata('shp',coastline,'dem',dem,...
                    'bbox',bbox,'h0',min_el);           
fh1 = edgefx('geodata',gdat1,...
            'fs',R,'wl',wl,'max_el',max_el,...
            'slp',slp,'dt',dt,'g',grade); 
mshopts = meshgen('ef',fh1,'bou',gdat1,'nscreen',10,'plot_on',1,'proj','stereo');
mshopts = mshopts.build; 
m = mshopts.grd; % get out the msh object    
m = interp(m,gdat1,'nan','fill','mindepth',1);

visualization of the m,

When I run write(m,'Global','ww3') , it shows an error in the function writeww3, this is because the variable opedat may be empty, so I modified part of the code of writeww3, (is this a bug?)

if(isempty(opedat))
    fprintf(fid,'%d\n', length(EToV(:,1)));
    m=0;
else
    fprintf(fid,'%d\n', length(EToV(:,1))+opedat.neta);
    m=0;
    for i=1:opedat.nope
        for j=1:opedat.nvdll(i)
            m=m+1;
            fprintf(fid,['%d %s %d %s %d %s %d %s %d %s %d\n'], m,'',15,'',2,'',0,'',0,'',opedat.nbdv(j,i));
        end
    end
end

I use the obtained mesh as input grid for WAVEWATCH3, the final output Netcdf file of WAVEWATCH3 shows the mesh is problematic,

This may be due to the fact that the generated grid surrounds the whole Earth. Steven R. Brus (https://gmd.copernicus.org/articles/14/2917/2021/) also points out the need to modify elements that straddle the −180–+180∘ boundary.

Describe the solution you'd like
Is it possible to remove the grid elements around -180 (or 180)?

Describe alternatives you've considered

When the grid range is determined as [-179.9,179.9;-89.9,89.9], a similar situation as above is encountered.
When the grid range is determined as [-179,179;-78,78], the final Netcdf shows the mesh is good.
when the grid range is determined as [-179,179;-89,89], the Mediterranean Sea is not taken into account.
Replace the projection from stereo to Robinson.
Due to the following code in meshgen.m, you need to set h0 to very small for global meshes (stereo does not need to set h0 to very small to get global meshes).

% kjr Oct 2018 use outer coarsest box for
% multiscale meshing
lon_mi = obj.bbox{1}(1,1)-obj.h0(1)/1110;
lon_ma = obj.bbox{1}(1,2)+obj.h0(1)/1110;
lat_mi = obj.bbox{1}(2,1)-obj.h0(1)/1110;
lat_ma = obj.bbox{1}(2,2)+obj.h0(1)/1110;

Additional context
This is the first time I submit an issue on github, there may be problems with the format of the question, and I'm looking forward to getting a reply, thanks~~

Mackay, Edward B.L. (2009) Wave energy resource assessment. University of Southampton, School of Civil Engineering and the Environment, Doctoral Thesis, 341pp.

1

2

First of all, thank you so much for developing this tool, it is very useful!

Is your feature request related to a problem? Please describe.

I want to create a globalized unstructured grid for WAVEWATCH3 to run, and the code to generate the grid is as follows (using the Example_7 program as a template):

min_el    = 20e3;  	             % pdf 3.1, minimum resolution in meters.
bbox      = [-180 180;-89 90];   % pdf 2.4, lon min lon max; lat min lat max
max_el    = 120e3; 		         % pdf 3.1, maximum resolution in meters. 
wl        = 30;                  % pdf 3.4, 30 elements resolve M2 wavelength.
dt        = 0;                   % pdf 4.2, Only reduces res away from coast
grade     = 0.35;                % pdf 4.1, mesh grade in decimal percent. 
R         = 3; 			         % pdf 3.3.2, Number of elements to resolve feature.
slp       = 10;                  % pdf 3.5, slope of 10
dem       = 'SRTM15+.nc';        % 
coastline = {'GSHHS_f_L1','GSHHS_f_L6'};
gdat1 = geodata('shp',coastline,'dem',dem,...
                    'bbox',bbox,'h0',min_el);           
fh1 = edgefx('geodata',gdat1,...
            'fs',R,'wl',wl,'max_el',max_el,...
            'slp',slp,'dt',dt,'g',grade); 
mshopts = meshgen('ef',fh1,'bou',gdat1,'nscreen',10,'plot_on',1,'proj','stereo');
mshopts = mshopts.build; 
m = mshopts.grd; % get out the msh object    
m = interp(m,gdat1,'nan','fill','mindepth',1);

visualization of the m,

When I run write(m,'Global','ww3') , it shows an error in the function writeww3, this is because the variable opedat may be empty, so I modified part of the code of writeww3, (is this a bug?)

if(isempty(opedat))
    fprintf(fid,'%d\n', length(EToV(:,1)));
    m=0;
else
    fprintf(fid,'%d\n', length(EToV(:,1))+opedat.neta);
    m=0;
    for i=1:opedat.nope
        for j=1:opedat.nvdll(i)
            m=m+1;
            fprintf(fid,['%d %s %d %s %d %s %d %s %d %s %d\n'], m,'',15,'',2,'',0,'',0,'',opedat.nbdv(j,i));
        end
    end
end

I use the obtained mesh as input grid for WAVEWATCH3, the final output Netcdf file of WAVEWATCH3 shows the mesh is problematic,

Describe the solution you'd like
Is it possible to remove the grid elements around -180 (or 180)?

Describe alternatives you've considered

When the grid range is determined as [-179.9,179.9;-89.9,89.9], a similar situation as above is encountered.
When the grid range is determined as [-179,179;-78,78], the final Netcdf shows the mesh is good.
when the grid range is determined as [-179,179;-89,89], the Mediterranean Sea is not taken into account.
Replace the projection from stereo to Robinson.
Due to the following code in meshgen.m, you need to set h0 to very small for global meshes (stereo does not need to set h0 to very small to get global meshes).

% kjr Oct 2018 use outer coarsest box for
% multiscale meshing
lon_mi = obj.bbox{1}(1,1)-obj.h0(1)/1110;
lon_ma = obj.bbox{1}(1,2)+obj.h0(1)/1110;
lat_mi = obj.bbox{1}(2,1)-obj.h0(1)/1110;
lat_ma = obj.bbox{1}(2,2)+obj.h0(1)/1110;

Additional context
This is the first time I submit an issue on github, there may be problems with the format of the question, and I'm looking forward to getting a reply, thanks~~

liu-jincan / study-fpm Goto Github PK

study-fpm's People

Contributors

Watchers

study-fpm's Issues

9

6

4

ss

Global mesh

Mackay, Edward B.L. (2009) Wave energy resource assessment. University of Southampton, School of Civil Engineering and the Environment, Doctoral Thesis, 341pp.

1

2

5

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent