A TEST OF AERIAL PHOTO CLASSIFICATIONS IN 



FOREST MANAGEMENT-VOLUME INVENTORIES 



INTRODUCTION 



The growing interest in combined forest management -volume inventories 1 and continuing 

 developments in aerial photo volume estimating justify a critical look at the forest classifica- 

 tions now used on aerial photos. Most of these classification schemes- -whether based on spe- 

 cies, stand size, age, stocking, volume, or site- -are those which evolved from experience in 

 ground surveys. Conceived primarily as map strata, their basic definitions remain those 

 developed before aerial photos or statistics were considered foresters' working tools. Few of 

 these basic criteria are directly measurable on aerial photos . 



These standard classifications, although traditionally recommended in forestry texts, are 

 often inefficient when used on aerial photos. During the past 20 years, many forest photo in- 

 terpreters have attempted to take the type, stand- size, and density classes used in timber 

 cruising and redefine them in terms recognizable on aerial photos . Some schemes have been 

 based on photo interpretation keys which made use of aerial or ground stereograms; others 

 relied wholly upon written descriptive material. The use of photo interpretation to recognize 

 these familiar but often arbitrary field classifications is subjective to a large extent. This is 

 an important weakness. 



The use of aerial photos in a survey usually implies double sampling. In double sampling 

 a large number of plots are classified on the photos and then a subsample of these is measured 

 on the ground to obtain mean volumes . The subsample is subsequently reclassified on the 

 ground in order to adjust the proportions of photo plots on which area estimates are based. 

 Although photo and ground classifications have the same designation, different tree or stand 

 features are measured. For example, in classifying by stand size, tree diameters are measured 

 on the ground, whereas the photo interpreter can only measure height and crown diameter. Al- 

 though these criteria are related, different classifications often result. These differences are 

 considered photo interpretation errors and are used to adjust the photo estimates of area. As 

 errors, they are assumed to be reflections on the effectiveness of photo stratification. 



This roundabout procedure, wherein photo interpretation is expected to classify plots into 

 essentially ground-defined strata, is justified on the assumption that the traditional ground 

 classifications are the most efficient approach to volume estimates; but, are they? 



Since the primary purpose of stratification in timber cruising has always been to obtain a 

 good volume estimate, the most effective photo scheme to accomplish this will be one whose 

 criteria are most closely related to ground volume. If the objectives also include an estimate 

 of acres needing cultural or other management treatment, the best scheme may be one whose 

 criteria are also related to these. Before traditional ground classification schemes are 

 recommended for photo use, they should be tested. 



Inventories made to assess cultural and improvement opportunities as well as area, 

 volume, growth, and mortality in forest stands. 



