Films and filters for identification of hardwood species in Scotland

This work is carried out to compare the merits of four different film-filter combinations for the identification of hardwood species in mixed forests in Scotland. The method of comparison is adopted to suit its application in tropical developing countries, for the identification of hardwood species in natural forests. Previous work on the use of aerial photography in forestry, with particular reference to the identification of species is reviewed, drawing conclusions for the need of separate tests, for the use of film-filter combination in different areas and conditions. Chapter one, deals with the characteristics of different films and filters which are extensively used in forest inventories. The suitability of different combinations of films and filters for surveying particular types of forest or species are also discussed. The detailed characteristics of four films (panchromatic, infrared, full colour and false colour) used in this investigation are given. In chapter two, the two forest areas selected for investigation, Dalmeny Estate (West Lothian) and Brechfa forests (Carmarthenshire), are described. The method of procuring aerial photographs, with their flight design, exposure data and date of photography are given. Chapter three, explains the method used in this experiment to compare the relative valves of the four film-filter combinations for species identification. Twenty sample trees are selected for interpretation from each of 20 sample plots. The sample trees in each plot are then grouped according to the appearance of their crowns on each of the sets of photographs. In order to minimise learning errors the order of using the different sets of photographs is randomised. The identity of the sample trees in each group is established by field survey. Chapter four, deals with the results and statistical analysis. The grouping achieved on each of the film-filter combinations is tested for two -types of error: (a) Omission error, which is the failure to group individuals of same species, in a single group. (b) Commission error, which is the confounding of individuals of different species. Analysis of Variance test is used to find out whether the results between the films are statistically significant. Student's 't' test is used to find out the real difference between the results of two sets of film-filter combinations. These results are discussed in chapter five and the following conclusions are drawn:- 1. Colour photographs are superior to black and white photographs for the separation of hardwood species in a mixed woodland. The results are statistically significant. 2. The least number of points scored by the false colour transparency indicates a definite trend of its superiority over the full colour. 3. Attention is drawn to the conditions required. for successful hardwood species :identification. The following requirements are not fulfilled in the present exercise. It is assumed that these shortcomings contribute to the level of error: (a) The preferred time of photography is mid-summer. (b) Scale of aerial photography should be larger than 1:10,000. (c) Use of anti-vignetting filters for full colour photography. (d) Good quality prints. 4. Errors due to incorrect transfer of sample trees from one photograph to the other is checked and found to be insignificant. 5. It is considered that the useful information was obtained with each of the films. Aerial photographic interpretation and ground work are deliberately separated in order to test the valve of each film-filter combination. In normal practice, the aerial photo-interpretation and ground work are closely integrated and a higher degree of success is thus obtained. 6. Further research should be carried out to compare the merits of only the full colour and false colour films for the identification of hardwood species in Scotland. The aerial photographs should be obtained simultaneously in mid-summer (July-August) at a scale larger than 1:10,000, using an anti-vignetting filter over the lens for full colour photography. Good quality full colour prints should be obtained by trial and error for comparison. The use of 23cm. square films for this type of investigation seems to be expensive at present, so 70mm colour aerial films could be conveniently used.