Background
This management scheme aims at a situation characterized by:
- parallel simulation of independent stands with a size of one hectare each
- a specific yield table management is defined for each hectare
- the management simply controls the stem numbers
Solution
Thanks to the mighty javascript engine, a script based solution is rather straightforward. The solution consists of a Javascript-part and a CSV-based textfile containing the management definition.
Javascript code
A brief explanation of the function is provided below. But here is the code itself:
/* Javascriptfile for iLand management. */
// global variables
var mgmt_list = {}
var mgmt_loaded = false;
/* Load the management descriptions from a CSV file.
Each line consists of a location (x/y), a year and a number of trees that should remain after management.
The data is stored as properties of an object (mgmt_list). */
function load()
{
mgmt_list = {}; // clear array
//var csvFile = new CSVFile; // (old Qt4 style)
var csvFile = Factory.newCSVFile(Globals.path('scripts/parallel_list.csv'));
var key, value;
for (i=0;i<csvFile.rowCount;i++) {
// create a combined key
key = csvFile.value(i,0) + '_' + csvFile.value(i,1) ; // ruindex_year
value = csvFile.value(i,2); // the remaining trees
mgmt_list[key] = value;
print (key); // just for debugging
}
}
// helping function to query the "database" of management operations
// returns <undefined> if no entry is inside the list
function queryMgmtList(ruindex, year)
{
var key = ruindex + '_' + year;
return mgmt_list[key];
}
/* callback function called by iLand.
executes management for each resource unit.*/
function manage(year)
{
if (mgmt_loaded == false) {
print("1st call - loading management...");
load();
mgmt_loaded = true;
print("loading management finished....");
}
print("executing management - year " + year);
var remains;
for (var i=0;i<Globals.resourceUnitCount;i++) {
remains = queryMgmtList(i, year);
if (remains!=undefined) {
// do some management
var tree_count = management.loadResourceUnit(i); // load all trees of resouce unit with index i
print("do management for resource unit " + i + ". Trees on RU: " + tree_count +". Trees to remain:" + remains);
// calculate how many trees are to be removed
var to_remove = tree_count - remains;
if (to_remove > 0) {
// killPct(percentile_from, percentile_to, count)
// killPct(0,100,x) -> kill x trees randomly
// you can apply some sorting:
// management.sort('dbh'); // sort by dbh ascending
// management.killPct(0,33,x); // -> remove x from lowest third
management.killPct(0,100,to_remove);
print("killed " + to_remove+" trees from resource unit " + i + ".");
}
}
}
}
Loading the data
The function load() loads the content from a tabular textfile and stores it in a simple datastructure. It uses the CSVFile object which is actually a binding to a C++ class of iLand.
The input table should have three columns: the index-number of the resource unit, a year and the number of trees that should remain after the management operation. Note that the resource unit index is 0-based, and that the first simulation year is year 1.
A sample file looks like:
## sample file for the management definition ## use either tab or ; as delimiter ## ***************************************** resourceUnitIndex year remaining 0 1 5000 0 10 4000 0 20 2500 0 30 1800 0 40 900 0 50 500 1 1 250 1 10 200 1 20 150 1 30 100 1 40 50 1 50 10 2 1 5000 2 10 3456 2 20 2333 2 30 1789 2 40 852 2 50 400
The management
The manage() function deals with the actual management. If called the first time, load() is called and the data is setup.
The real work starts with looping through all resource units of the system (using the model variable Globals.resourceUnitCount). For each unit the function checks if a entry is present in the mgmt_list object.
If no data is present for the combination of year and resource unit, nothing happens.
Otherwise, the actual management is performed: the current number of trees of the resource unit is compared with the target number; the surplus number of trees is then removed.
;){kind=link}