LISFLOOD-FP and MATLAB

From SourceWiki
Revision as of 17:48, 18 February 2009 by Tim-fewtrell (talk | contribs)
Jump to navigation Jump to search

This page contains various Matlab functions that may be useful for analysing results from the LISFLOOD-FP model.

File import and export

Importing ascii raster files

This function will import an ascii raster file using the string filename. To use it, copy and past text into a textfile called ascii_reader.m

function [dem, ncols, nrows, xllcorner, yllcorner, cellsize] = ascii_reader (filename) 

% [DEM, ncols, nrows, xllcorner, yllcorner, cellsize] = ascii_reader (filename) 

% This function reads arc .asc files
% requires filename string as input

% j neal
% 18/6/07
%% read an ascii file
fin = fopen(filename,'r');
A = fscanf(fin,'%s',1); ncols = fscanf(fin,'%f',1);           %#ok<NASGU>
A = fscanf(fin,'%s',1); nrows = fscanf(fin,'%f',1);           %#ok<NASGU>
A = fscanf(fin,'%s',1); xllcorner = fscanf(fin,'%f',1);       %#ok<NASGU>
A = fscanf(fin,'%s',1); yllcorner = fscanf(fin,'%f',1);       %#ok<NASGU>
A = fscanf(fin,'%s',1); cellsize = fscanf(fin,'%f',1);        %#ok<NASGU>
A = fscanf(fin,'%s',1); nodata = fscanf(fin,'%f',1);          %#ok<NASGU>
dem = fscanf(fin,'%f',[ncols, nrows]);
dem = dem';
fclose('all');

You can view the imported file using the command:

imagesc(dem);

Exporting ascii raster files

This function will create an ascii raster file called filename from the 2D matrix DEM. To use it, copy and past text into a textfile called ascii_write.m

function ascii_write (filename, DEM, xllcorner, yllcorner, cellsize, nodata)

% ascii_write (filename, DEM, xllcorner, yllcorner, cellsize, nodata)
% only filename (string) and DEM (2D matrix) are essential
% this function writes ascii raster files for arc
% j neal
% 6/3/2008

if nargin < 2, 
    error('Requires at least two input arguments'); 
end
if nargin < 4, 
    xllcorner = 0;
    yllcorner = 0;
end
if nargin < 5,
    cellsize = 1;
end
if nargin < 6
    nodata = -9999;
end
A = size(DEM);
fout = fopen (filename,'w');
fprintf(fout, 'ncols         %.0f\n', A(2));  
fprintf(fout, 'nrows         %.0f\n', A(1));
fprintf(fout, 'xllcorner     %f\n', xllcorner);
fprintf(fout, 'yllcorner     %f\n', yllcorner);
fprintf(fout, 'cellsize      %f\n', cellsize);
fprintf(fout, 'NODATA_value  %f\n', nodata); 
for ii = 1:A(1)
    B = DEM(ii,:);
    fprintf(fout, '%1.3f ', B);
    fprintf(fout, '\n');
end
fclose('all');

Importing .profile files

This function imports .profile files

function [data, M] = profileread (resroot, numfiles, t, string1, string2)

% This function reads LISFLOOD-FP .profile files
%
% DATA = PROFILEREAD (RESROOT)
%   where RESROOT is a string containing the resroot from the LISFLOOD-FP
%   .par file. Absolute path names can also be used e.g. 'C:\LISFLOOD\resroot'
%   DATA is a structure containing .header {cell aray}
%                                  .segmentN (numerical array for segment N)
%
% DATA = PROFILEREAD (RESROOT, NUMFILES)
%   NUMFILES = number of files to read. Reads m files assuming that the files 
%   are numbered such that resroot-000m.profile. 
%   DATA(m) is a structure containing m files
% 
% DATA = PROFILEREAD (RESROOT, NUMFILES, 1)
%   specifies the profiles were exported a user specified time such that
%   resroot-000x-t.profile
%
% [DATA, M] = PROFILEREAD (RESROOT, NUMFILES, 1, string1, string2)
%   instructs the function to plot string1 or sting1 against string2... these
%   strings muct match the headernames in the .profile file e.g. 'Flow'
%   returns movieframes in the structure M. Also plots the data.
%
% Note: LISFLOOD-FP will name multiple files 0 to m-1 but DATA(m) will count from 1 to m
%
% j.neal
% 12/5/2008
%%
if nargin < 1, 
   error('Requires filename'); 
end
if nargin < 2,
   % only one file
   numfiles = 0;
end
if nargin < 2,
   t = 0;
end
if nargin < 4,
   % no plotting required
   plotting = 0;
   M = 0;
elseif nargin < 5,
   % use plotter
   plotting = 1;
elseif nargin < 6,
   plotting = 2;
end
%% Arrange data read
if t == 1
   fileex = '-t.profile';
else
   fileex = '.profile';
end
if numfiles == 0
   % read in a single .profile file
   filename = [resroot,fileex];
   data = read_profile (filename);
elseif numfiles < 9999
   % read a series of .profile files
   for i = 1:numfiles
       ii = i-1;
       if ii < 10
           resroot2 = [resroot,'-000',num2str(ii),fileex];
       elseif ii < 100
           resroot2 = [resroot,'-00',num2str(ii),fileex];
       elseif ii < 1000
           resroot2 = [resroot,'-0',num2str(ii),fileex];
       else
           resroot2 = [resroot,'-',num2str(ii),fileex];
       end
       %returnes structure with format data(i).segmentX... also includes .header 
       [data(i), seg_num] = read_profile (resroot2); %#ok<AGROW>
   end
else
   error ('Too many files or incorrect input to numfiles'); 
end
%% Plotting under development
% account for 0 seg num
seg_num = seg_num +1;
% if plotting == 0 % do nothing... no plots
if plotting == 1
   % find the column you want to plot
   I = strmatch(string1 , data(1).header);
   % plot a single variable
   A = size(data);
   for i = 1:A(2)
       figure(i);
       for j = 1:seg_num
           k = j-1;
           subplot(1,seg_num,j);
           eval(['plot(data(i).segment',num2str(k),'(:,',num2str(I),'));']);
           tstring = ['Plot of ',resroot,num2str(i),' segment ',num2str(k)];  
           title(tstring);
           % get movie frame
           M(i,j) = getframe(gcf); %#ok<AGROW>
       end
   end
elseif plotting == 2
   % find the columns you want to plot
   I = strmatch(string1 , data(1).header);
   I2 = strmatch(string2 , data(1).header);
   A = size(data);
   for i = 1:A(2)
       figure(i);
       for j = 1:seg_num
           k = j-1;
           subplot(1,seg_num,j);
           eval(['plot(data(i).segment',num2str(k),'(:,',num2str(I),'),data(i).segment',num2str(k),'(:,',num2str(I2),'));']);
           % get movie frame
           tstring = ['Plot of ',resroot,num2str(i),' segment ',num2str(k)];  
           title(tstring);
           M(i,j) = getframe(gcf); %#ok<AGROW>
       end
   end
end
%% read .profile file
function [data, seg_num] = read_profile (filename) 
fin = fopen(filename,'r');
more_segs = 1;
% read 'Channel_segment:' from file
temp = fscanf(fin,'%s',1); %#ok<NASGU>
while more_segs == 1
   % read filename
   seg_num = fscanf(fin,'%f',1);
   for i = 1:11
       data.header{i} = fscanf(fin,'%s',1);
   end
   testline = 1;
   j = 0;
   while testline == 1
       j = j+1;
       % reads the first bit of the line as a string
       A = fscanf(fin, '%s',1);
       % this string is either a numer of a new channel sgment of the end
       % of the file
       % if end of file
       ST = feof (fin);
       if ST == 1
           % reached the end of the file exit both while loops
           testline = 0;
           more_segs = 0;
       else
           TF = strcmp('Channel_segment:', A);
           if TF == 1
               % new segment has been found
               testline = 0;
           else
               % read rest for line into data
               B(j,1) = str2double(A); %#ok<AGROW>
               for k = 2:11
                   B(j,k) = fscanf(fin, '%f',1); %#ok<AGROW>
                end
            end
        end
    end
    eval(['data.segment',num2str(seg_num),' = B;']);
    clear B
end
fclose('all');

Writing parameter files

This function can be used to write many LISFLOOD-FP parameter files with N different channel and floodplain roughness, using your own N-by-2 matrix called roughness. You will need to edit the code for your application by commenting unwanted lines with a percentage symbol.

function prm_writer (roughness) 

% This function writes LISFLOOD-FP parameter files numbered 1 to n using
% the roughness values in the n-by-2 matrix roughness. yuo will need to
% edit the code for your application.

[num_sims] = size(roughness);
for i = 1:num_sims(1)
a = ['mymodel', num2str(i),'.par']; b = num2str(i);
fout = fopen(a,'w');
fprintf(fout,'DEMfile                  mydem.dem.ascii\n');
fprintf(fout,'resroot                  myres%s\n',b);
fprintf(fout,'dirroot                  mydir\n');
fprintf(fout,'sim_time                 245000.0\n');
fprintf(fout,'initial_tstep            10.0\n');
fprintf(fout,'massint                  300.0\n');
fprintf(fout,'saveint                  10000.0\n');
%fprintf(fout,'overpass                 135000.0\n');
fprintf(fout,'fpfric                   %1.3f\n',roughness(i,2));
fprintf(fout,'nch                      %1.3f\n',roughness(i,1));
%fprintf(fout,'manningfile              mymanfile%s.n.ascii\n',b);
fprintf(fout,'riverfile                myriver.river\n');
fprintf(fout,'bdyfile                  mybdy.bdy\n');
fprintf(fout,'stagefile                mystage.stage\n');
fprintf(fout,'bcifile                  mybci.bci\n');
%fprintf(fout,'startfile                mystart.old\n');
%fprintf(fout,'checkpoint               2\n');
%fprintf(fout,'checkfile                mycheck.chkpnt\n');
fprintf(fout,'elevoff\n');
fprintf(fout,'qoutput\n');
fprintf(fout,'diffusive\n');
fclose('all');
end

Plotting flow vectors from .Qx and .Qy files

This function will plot flow vectors from a .Qx and .Qy over a dem using the matlab functions image and quiver

function lisflood_flow_plotter (qxfile, qyfile, demfile, north, east, south, west, scaling)

% This function plots flow vectors from LISFLOOD_FP Qx and QY files
%
% lisflood_flow_plotter (qxfile, qyfile, demfile);
%
% these input variables are strings containing the relative of full
% pathnames of the .Qx .Qy and ascii dem files to be plotted.
%
% lisflood_flow_plotter (qxfile, qyfile, demfile, N, E, S, W);
%
% N, E, W and S are optional they specify the northernmost, southeernmost
% ect cell to be displayed. this can be used to crop the domain.
%
% lisflood_flow_plotter (qxfile, qyfile, demfile, N, E, S, W, scaling);
% 
% scaling is an optional term used to manually scale the vector lengthes
% drawn by quiver (the arrow plotting function).
%
% J Neal
% 20/02/2008 

%% decide if crop is required and give imput error message
if nargin < 3, 
    error('Requires at least three input arguments'); 
end
if nargin < 7, 
    crop = 0;
else
    crop = 1;
end
if nargin < 7,
    scaling = 1;
end 
% Note don't do this for large areas ?
%% Task 1: read ascii files
% read Qx file
[QX, ncolsqx, nrowsqx, xllcornerqx, yllcornerqx, cellsizeqx] = read_file (qxfile); %#ok<NASGU>
% read Qy file
[QY, ncolsqy, nrowsqy, xllcornerqy, yllcornerqy, cellsizeqy] = read_file (qyfile); %#ok<NASGU>
% read DEM
[DEM, ncolsdem, nrowsdem, xllcornerdem, yllcornerdem, cellsizedem] = read_file (demfile);
%% crop data if required
if crop == 1;
    DEM = DEM(north:south,west:east);
    QX = QX(north:south,west:east+1);
    QY = QY(north:south+1,west:east);
    xllcornerdem = xllcornerdem + west*cellsizedem - cellsizedem;
    yllcornerdem = yllcornerdem + (nrowsdem - south) * cellsizedem;
    [nrowsqx, ncolsqx] = size(QX); [nrowsqy, ncolsqy] = size(QY); [nrowsdem, ncolsdem] = size(DEM);
end
%% Taks 2: Plot dem and generate figure
figure1 = figure('PaperSize',[20.98 29.68],'Color',[1 1 1]);
colormap('gray');
% Create axes
axes1 = axes('Parent',figure1,'YTickLabel',{yllcornerdem + nrowsdem*cellsizedem ,yllcornerdem},...
    'YTick',[0.5,nrowsdem+0.5],...
    'YDir','reverse',...
    'XTickLabel',{xllcornerdem,xllcornerdem+ncolsdem*cellsizedem},...
    'XTick',[0.5, ncolsdem + 0.5],...
    'Layer','top',...
    'DataAspectRatio',[1 1 1]...
    %,'Clim',[13 30]... % uncomment to set colourmap limits manually
    );

box('on');
hold('all');
image(DEM,'Parent',axes1,'CDataMapping','scaled');
axis('image');
xlabel('Easting (m)');
ylabel('Northing (m)');
%% Task 3: calculate flow vectors
% pre allocate memory for loop... it will run faster this way
xlocs = zeros(ncolsqx*nrowsqx+ncolsqy*nrowsqy,1);
ylocs = xlocs;
U = xlocs;
V = xlocs;
%work out X and Y locations
for i = 1:nrowsqx
    for j = 1:ncolsqx
        xlocs((i-1)*ncolsqx+j,1) = j - 0.5;
        ylocs((i-1)*ncolsqx+j,1) = i;
        U((i-1)*ncolsqx+j,1) = QX(i,j);
    end
end
for i = 1:nrowsqy
    for j = 1:ncolsqy
        xlocs(ncolsqx*nrowsqx + (i-1)*ncolsqy +j,1) = j;
        ylocs(ncolsqx*nrowsqx + (i-1)*ncolsqy +j,1) = i - 0.5;
        V(ncolsqx*nrowsqx+(i-1)*ncolsqy+j,1) = QY(i,j);
    end
end
% work out number of nonzeros elements
num_flows = nnz(U+V);
xlocs2 = zeros(num_flows,1);
ylocs2 = xlocs2;
U2 = xlocs2;
V2 = xlocs2;
% remover zero flow locations
k = 1;
for i = 1:ncolsqx*nrowsqx+ncolsqy*nrowsqy
    if (U(i,1) == 0) && (V(i,1) == 0)
        % do nothing
    else
        xlocs2(k,1) = xlocs(i,1);
        ylocs2(k,1) = ylocs(i,1);
        U2(k,1) = U(i,1);
        V2(k,1) = V(i,1);
        k = k+1;
    end
end
% check numbers are correct
if k == num_flows + 1
    % all is ok
else
    disp('Problem with flows');
end
%% Task 4: overlay flow vectors
disp('Plotting data... if this is taking a long time you may need fewer locations');
quiver (xlocs2,ylocs2,U2,V2, scaling);
% quiver (xloc2,yloc2,U2,V2,'Parent',figure1);
disp('Done');
%% function to read LISFLOOD ascii files using filename
function [OUT, ncols, nrows, xllcorner, yllcorner, cellsize] = read_file (filename)
% read an ascii file
fin = fopen(filename,'r');
A = fscanf(fin,'%s',1); ncols = fscanf(fin,'%f',1);           %#ok<NASGU>
A = fscanf(fin,'%s',1); nrows = fscanf(fin,'%f',1);           %#ok<NASGU>
A = fscanf(fin,'%s',1); xllcorner = fscanf(fin,'%f',1);       %#ok<NASGU>
A = fscanf(fin,'%s',1); yllcorner = fscanf(fin,'%f',1);       %#ok<NASGU>
A = fscanf(fin,'%s',1); cellsize = fscanf(fin,'%f',1);        %#ok<NASGU>
A = fscanf(fin,'%s',1); nodata = fscanf(fin,'%f',1);          %#ok<NASGU>
OUT = fscanf(fin,'%f',[ncols, nrows]);
OUT = OUT';
fclose('all');

Making movies from .wd files

This function will convert a series of .wd .Qx or .Qy file into a .avi movie with the same filename

function LISFLOOD_mov (resroot,fileex,vartype,num_snaps,snapint,dem,demCS,depthCS,framesPS,movQ) 

% LISFLOOD_mov generates movies of LISFLOOD-FP output
%
% LISFLOOD_mov (resroot,fileex,vartype,num_snaps,snapint,dem,demCS,depthCS,framesPS,movQ)
% 
% where 
% resroot is the LISFLOOD resroot (relative to m file or absolute)(string)
% fileex is the file extension '.wd' (can be chaned to plot WSE and flows)
% vartype is the name of the variable being plotted e.g. 'Depth'
% num_snaps is the number of snapshot times
% snapint is the time interval between each LISFLOOD-FP snapshot (seconds)
% dem is the name of the demfile (relative or absolute)
% demCS is the range of dem hights to be plotted e.g. demCS = [0,20];
% depthCS is the range of depth values to be plotted e.g. depthCS = [0,10];
% framesPS number of frames per second (each plot is a simgle frame)
% movQ is the movie quality between 1 and 100 (100 is best)
% 
% 8/11/2007. J Neal.
%
% EXAMPLE:
% LISFLOOD_mov ('C:\Experiment1\res22','.wd','Water depth',24,10000,...
% 'C:\Experiment1\dem10m.dem.ascii',[0,60], [0,10],1,100);
%% Movie maker
% read dem filenam is filename and 1 specifies the 0.00 should not be NaN's
[DEM, ncols, nrows, xllcorner, yllcorner, cellsize] = read_file (dem); %#ok<NASGU>
% generate empty .avi file
% A = [resroot,'.avi'];
% mov = avifile(A);
% loop through snapshots
for i = 1:num_snaps
    if i < 10
        resfile = [resroot,'-000',num2str(i),fileex];
    elseif i < 100
        resfile = [resroot,'-00',num2str(i),fileex];
    elseif i < 1000
        resfile = [resroot,'-0',num2str(i),fileex];
    else
        resfile = [resroot,'-',num2str(i),fileex];
    end
    % read in deoth file for resfile i (2 specifies the 0.00 should be NaN
    [DEPTH, ncols, nrows, xllcorner, yllcorner, cellsize] = read_file (resfile);
    % plot data as figure 1
    plotter (DEPTH, DEM, nrows, ncols, cellsize, xllcorner, yllcorner,demCS,depthCS,snapint,i,vartype);
    % get movie frame
    M(i) = getframe(gcf); %#ok<AGROW>
    hold off;
%     mov = addframe(mov,M(i));
    % close the figure
    close all
end 
% % close .avi file
% mov = close(mov);
% generate empty .avi file
disp('Generating .avi movie file');
A = [resroot,fileex,'.avi'];
movie2avi(M,A,'fps',framesPS,'quality',movQ);
disp(A);
%% Read LISFLOOD ascii files using filename
function [OUT, ncols, nrows, xllcorner, yllcorner, cellsize] = read_file (filename)
% read an ascii file
fin = fopen(filename,'r');
A = fscanf(fin,'%s',1); ncols = fscanf(fin,'%f',1);           %#ok<NASGU>
A = fscanf(fin,'%s',1); nrows = fscanf(fin,'%f',1);           %#ok<NASGU>
A = fscanf(fin,'%s',1); xllcorner = fscanf(fin,'%f',1);       %#ok<NASGU>
A = fscanf(fin,'%s',1); yllcorner = fscanf(fin,'%f',1);       %#ok<NASGU>
A = fscanf(fin,'%s',1); cellsize = fscanf(fin,'%f',1);        %#ok<NASGU>
A = fscanf(fin,'%s',1); nodata = fscanf(fin,'%f',1);          %#ok<NASGU>
OUT = fscanf(fin,'%f',[ncols, nrows]);
OUT = OUT';
fclose('all');
% convert nodata to NaN;
for ii = 1:nrows
    for jj = 1:ncols
        if OUT(ii,jj) == nodata
            OUT(ii,jj) = NaN;
        elseif OUT(ii,jj) == -99.000
            % lisflood uses -99.000 as nodat so will also remove these
            OUT(ii,jj) = NaN;
        else
            % do nothing
        end
    end
end
%% Plotter data for movie slide
function plotter (DEPTH, DEM, nrows, ncols, cellsize, xllcorner, yllcorner,demCS,depthCS,snapint,i,vartype)
% number of colors in colormap
num_colors = 64;
% DEM and variable color map
cmap = [gray(num_colors); jet(num_colors)];
% scale DEM and DEPTH data to fit with new colormap
demCS2 = demCS(2)-demCS(1);
dem_scalar = num_colors/demCS2;
DEM2 = DEM*dem_scalar;
depthCS2 = depthCS(2)-depthCS(1);
depth_scalar = num_colors/depthCS2;
DEPTH2 = (DEPTH*depth_scalar)+num_colors+1;
%Plot the DEM
figure1 = figure('PaperSize',[20.98 29.68],'Color',[1 1 1]);
colormap(cmap);
axes1 = axes('Parent',figure1,'YTickLabel',{yllcorner+(nrows*cellsize),yllcorner+cellsize},...
    'YTick', [1,nrows],...
    'YDir','reverse',...
    'XTickLabel',{xllcorner,xllcorner+(ncols*cellsize)-cellsize},...
    'XTick',[1,ncols],...
    'Layer','top');
box('on');
hold('all');
image(DEM2,'Parent',axes1);
axis('equal');axis('tight');
% Overlay the water depth (or other variable)
H = image(DEPTH2);
% code to make depth plot layer transparent
OP = ones(nrows,ncols);
for ii = 1:nrows
    for jj = 1:ncols
        if DEPTH(ii,jj) > 0
        else
            OP(ii,jj) = 0;
        end
    end
end
% make transparent. depth layer must be H
set(H,'AlphaData',OP);
% label colourbar
color_breaks = 5;
colorbreak = num_colors/color_breaks;
SCALE = zeros(color_breaks*2,1);
SCALE(1) = 0.0001;
for j = 1:color_breaks*2
    SCALE(j+1) = colorbreak*j;
end
SCALE(color_breaks+1) = num_colors+1;
VIS_SCALE = zeros(color_breaks*2,1);
VIS_SCALE(1) = demCS(1);
for j = 1:color_breaks-1
    VIS_SCALE(j+1) = ((demCS2/color_breaks)*j)+demCS(1);
end
VIS_SCALE(color_breaks+1) = depthCS(1);
for j = 1:color_breaks
    VIS_SCALE(j+1+color_breaks) = ((depthCS2/color_breaks)*j)+depthCS(1);
end
colorbar('YTick',SCALE,'YTickLabel',VIS_SCALE);
% Label axis
xlabel('BNG easting (m)');
ylabel('BNG northing (m)');
A = [vartype,' over DEM at time ',num2str((i*snapint)),'s'];
title(A);