Seither wurde ein drittes Experiment. in einem tropischen Land durchgeführt. 
Hauptziel war die Fähigkeiten des Radar Höhenmessers, genaue topographische Profile 
und auch gleichzeitig Profile des Kronendaches zu ergeben, festzustellen. Beide Profile 
werden auf Papierstreifen registriert. Sekundäres Ziel war, die Genauigkeit der 
Baumartenerkennung zu testen, 
Die erhaltenen Daten werden gegenwärtig noch analysiert, aber es hat sich 
schon gezeigt, daß auch die tropische Vegetation das Signal des Radars nicht beeinflußt. 
Genaue topographische Profile wurden auch von dicht bewaldeten Gebieten von der Luft 
gewonnen. Das Kronendachprofil erweist eine Korrelation mit Baumhöhe und 
Baumdichte auf. Auch die Baumartenerkennung erscheint vielversprechend, 
Introduction 
The Forest Management Institute has been developing a method to obtain 
forest inventory statistics based on tree measurements from very large-scale photo- 
graphs for over 15 years. The method is now operational and is being implemented in the 
inventory programs of some of the Canadian provinces. The methodology has recently 
been described by Aldred and Lowe (1978). Very briefly, aerial photographs at scales of 
about 1:1000 to 1:4000 are obtained over sample locations. A foliage-penetrating radar 
altimeter and a tilt-measuring device provide the elements for an analytical absolute 
orientation of the stereo models. Tree heights and often crown areas are photogram- 
metrically measured and entered into regression equations which provide estimates of 
tree volume and diameter. The traditional field plot becomes a photo plot, but the 
procedures are similar to those of conventional forest inventories, 
The potential of this methodology to provide useful information in tropical 
surveys was recognized early in the development. The first test in Guatemala (1968) 
demonstrated that the tropical forest posed a considerable challenge to instrumentation 
and interpreter. Instrumentation was modified and tested again in Surinam in 1973. 
Some of these results were presented at the 1976 IUFRO World Congress (Nielsen and 
Aldred 1976). A more recent summary was given by Sayn-Wittgenstein (1978). 
The Surinam trials led to several recommendations for improvement of the 
system and further testing. These recommendations were acted upon by GTW Resource 
Inventory Radar Ltd., the altimeter's manufacturing company, supported by a contract 
awarded by the Canadian Department of Supply and Services. 
The new experiments took place in Costa Rica and the test area chosen was 
the field station "La Selva" owned by the Organization of Tropical Studies. 
Objectives of the experiments 
l. To establish how well the signal from a recently modified version of the forestry 
radar altimeter could penetrate tropical forest vegetation and, in conjunction with 
a barometric sensor, produce a ground elevation profile. 
2. To test if added circuitry would enable the radar altimeter to respond to the forest 
canopy and produce, simultaneously to the ground profile, a profile of the stand 
canopy. The difference between the, ground and canopy profiles would yield stand 
height which would be the variable used to assess the effectiveness of the double 
trace feature. 
To explore further the possibility of tree species identification on large-scale