18.04.2013 Views

do Instituto Hidrográfico

do Instituto Hidrográfico

do Instituto Hidrográfico

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

70 ANAIS DO INSTITUTO HIDROGRÁFICO N.º 15<br />

this computational environment, assessing data in a<br />

point or along a ran<strong>do</strong>m contour. From this reason it<br />

was introduced as an output the number of points. If<br />

this number is 1, for any point selected in the area is<br />

provided the depth and as a comparison measure is<br />

given also the location and the value of the maximum<br />

depth in the field (figure 7). If this number is greater can<br />

be generated a path in the field from these points and<br />

can be visualized the bathymetric distribution along it.<br />

In figure 8 is illustrated such a path created by 5 points<br />

and the corresponding bathymetry. It is also available an<br />

enlarged view of the bathymetric distribution where can<br />

be seen in detail the variations along the lines which<br />

compose the created path. For generating the path divisions<br />

was used the solution of varying the steps length<br />

and number on each line of the path. The reference step<br />

will be considered constant for a given bathymetric<br />

configuration and its value will be the minimum of the<br />

grid steps:<br />

d st r = min(DXC1, DYC1). (1)<br />

The reason for this election was that even with a<br />

smaller step the resolution would be the same because<br />

all the interpolations are made anyway between the<br />

points of the grid. Each line is defined by its endpoints<br />

and the length of the j-line will be given by:<br />

Llj = E (x 2j – x lj) 2<br />

+ (y 2j – y lj) 2<br />

, (2)<br />

where (x lj, y lj) and respectively (x 2j, y 2j) are the line end<br />

points. The number of the points on the line will be:<br />

Npj = V (Llj / d st r), (3)<br />

where V is the MATLAB ceil function which rounds the<br />

argument to the nearest greater integer. The new step<br />

for the j-line will be:<br />

d st j = Llj / Npj. (4)<br />

With this step the computational process will<br />

advance in space from the first to the last point of the<br />

line. If the distance between the line endpoints is less<br />

than dst l than on that line this distance will be the step<br />

dst j . The procedure is reinitialized for every line of the<br />

given path.<br />

Reshaping the configuration of the field is another<br />

procedure available in this new computational environment.<br />

The global variables that control the process of<br />

reshaping are: XCO2- the x coordinate of the new origin,<br />

YCO2- the y coordinate of the new origin, XLENC2- the<br />

length of the new grid in x direction, YLENC2- the<br />

length of the new grid in y direction, MXC2- the number<br />

of meshes in x direction and MYC2- the number of<br />

meshes in y direction. To illustrate this process it was<br />

elected another area in the vicinity of the Pinheiro da<br />

Cruz beach. Figure 9 shows the initial grid before<br />

reshaping while in figure 10 is presented the reshaped<br />

grid. In figure 11 can be seen the map of the area and<br />

the location of the reshaped area.<br />

Finally defining the mathematical isolines of the<br />

bottom it is also a process that can be developed using<br />

this interface. In this respect was created a special operator<br />

denoted by:<br />

C = C (wd, k sh, XoX, YoY, ZoZ) (5)<br />

in which wd is the imposed water depth for which is<br />

generated the isoline, k sh is the advanced shore coefficient<br />

which gives the sense of the iteration, if k sh is<br />

greater than wd this will be the positive sense of x axis<br />

and if k sh is less than wd is reversed. For considering the<br />

iteration along the y axis it was used a similar operator<br />

denoted by C * . XoX, YoY and ZoZ are the matrices<br />

giving the x coordinate, y coordinate and respectively<br />

the depth in the grid points. From the way it was defined<br />

is obvious that the above operators can be applied both<br />

for the initial and the reshaped grids. Supposing that<br />

the lines of the grid are parallel with the y axis, which is<br />

also almost parallel to the shore, and the columns are<br />

parallel with the x axis, the algorithm developed find for<br />

any column the grid point with the closer value (of wd).<br />

Then interpolates between the two successive grid<br />

points on the column. In this way the isoline is built<br />

with a single point in any column and will have the<br />

same number of points as a regular line of the grid,<br />

which is a great advantage for the MATLAB environment<br />

where the dimension of vectors and matrices should<br />

correspond.<br />

If no value is found, means that the depth imposed<br />

for the isoline is out of range and in this case the value<br />

is zero if there is a coast side in the area or either zero<br />

or maximum if there isn’t. If the coast is parallel with<br />

the x axis then the iteration is made along the y axis and<br />

it is used the operator C * . In the figures 11 and respectively<br />

12 is presented the way in which can be generated<br />

the isolines in the initial and respectively the reshaped<br />

area, while in figure 13 could be seen the natural<br />

isolines of the bottom. A further option available is to<br />

design automatically an isomap, that is to set the colors<br />

and the distance, in terms of depths, between them as it<br />

showed in figure 14.<br />

Another facility is to work in parallel, simultaneous<br />

on two different win<strong>do</strong>ws, with the initial and with the<br />

reshaped grid for both the processes of data assessment<br />

and generating isolines. Although were used the same<br />

programs for the initial and the reshaped grid this was<br />

possible because the system works with two categories<br />

of global variables (that have the index 1 for the initial<br />

surface and the index 2 for the reshaped one). These are<br />

changed alternatively from one case to another.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!