Field test of canopy cover estimation by hemispherical photographs taken with a smartphone

Abstract Aim To test the Canopy Cover (CaCo) index for forest vegetation research estimation of canopy cover from hemispherical photographs, and introduce a new Android smartphone application for use of this index. Methods The original and modified CaCo index were evaluated using a data set of 234 hemispherical photographs taken in 78 plots in coniferous, mixed and broad‐leaved deciduous forests. The results of the CaCo analysis of these photographs were compared with expert field visual estimation of canopy cover of these plots. For each hemispherical photograph, several CaCo values were calculated based on the photograph being restricted to different degrees by artificial horizon masking. The CaCo index was also tested with respect to precision of canopy cover estimation and sensitivity to different photographic equipment, using a different set of 93 canopy photographs taken in mixed and coniferous forests. Calculation of the CaCo index was done with the newly developed GLAMA – Gap Light Analysis Mobile Application software. Results Linear regression showed a close relationship between the CaCo index and visually observed canopy cover data. A proposed calculation modification improved the stability of the CaCo index in cases in which no horizon masking was applied. The best fit, zero‐intercept and a regression slope close to1 were found in cases in which an artificial horizon mask that extended higher than 45° restricted the bottom part of the sky hemisphere. Low sensitivity of the CaCo index to type of photographic equipment used was shown. Conclusions The CaCo index is robust and can be used for precise canopy cover estimation, comparable to visual canopy cover estimation and unaffected by observer bias. Not only can it be used on already‐captured photographs, but the index can also be employed on smartphones to rapidly capture hemispherical photographs and immediately calculate their index values. This application is freely available on the Internet and can serve as a powerful research and educational tool that can not only calculate CaCo values, but also standardize forest canopy visual estimates.