Estimating GPS Signal Loss in a Natural Deciduous Forest Using Sky Photography

Understanding how Global Positioning System (GPS) signals are influenced by vegetation structure allows for the determination of how specific technologies might be affected in certain forest environments. This study presents three different models that predict signal loss in a natural deciduous forest using fisheye photography. Relationships between terrestrial-based hemispherical sky-oriented photo (HSOP) measurements and GPS signal-to-noise ratios (SNRs) are explored. ArcGIS is used for image processing of HSOPs to rapidly estimate canopy closure (CC) at particular angles from zenith in forested areas. The difference between the observed SNR of GPS L-band signals under forest canopies to those observed in the open determines signal loss. CC values at different zenith angles inside the forest during four seasons are used to model signal attenuation. This article presents a canopy closure predictive model (CCPM), a model that includes the CCPM and incorporates the difference between the CC value in any season minus the CC in the winter, and a model that includes a seasonal component. The three models presented in this article yield adjusted R2 values between 0.60 and 0.62 and root mean square error range of 3.21 to 3.28 dB