Potential of Geospatial Technologies in Mechanized Timber Harvesting Planning

Abstract

Mechanized timber harvesting involves various activities including; road planning and selection of harvesting systems and machineries. The emergence of geospatial technology (GSPT) i.e., geographical information system (GIS) and remote sensing in the recent decades, has been considered as the best tools to facilitate timber harvesting planning in plantation forests. GSPT provide accurate stand information enabling better decision-making and optimizing forest operations. This study was conducted at Sao hill Forest Plantation (SHFP) in Tanzania, with the objective of determining relative efficiency (RE) between geospatial approach (GSPA) and conventional approach (CA) on planning mechanized timber harvesting. 120 grapple skidder (GS) time study observations in 30 sample plots covering different elevation terrain ranges were studied in both approaches. Productivity and costs under the two approaches were estimated and modelled using generalized linear model (GLM) approach. To obtain large scale estimates of productivity and costs, Inverse Distance Weighted (IDW) interpolation approach was used. The results showed that, GSPA demonstrated higher productivity and lower unit skidding costs (i.e., 71.1m3/hr and 2.121USD/m3) compared to CA (i.e., 67.5m3/hr and 2.914USD/m3) respectively. Skidding distance and slope (p-value < 0.05) were significant predictors of the GS performance in both approaches. The pseudo R2 ranging from 58.1% to 64.3% under CA, and from 62.9% to 60.8% under GSPA. Likewise, relative root mean square error (RMSEr) for the models under CA ranged from 49.3% to 50.4% and 33.4% to 35.2% under GSPA. Generally, the results showed that, models under GSPA have better fits and accuracy, compared to CA. Furthermore, the GSPA provided a raster representation of productivity and costs over the entire study area. Moreover, computed RE values (i.e., 1.18 and 6.17) indicated that parameter estimates for the GS productivity and costs were more precise in geospatial models (GSPM) compared to conventional models (CM). These findings highlight the potential of GSPT for an efficient large scale timber harvesting planning, by considering terrain constraints.