Combining auxiliary variables and ground data of forest parameters using the model-based approach has become a frequently used methodology to produce synthetic estimates for small areas. These small areas arise as it may not be financially feasible to take ground measurements or some areas may be inaccessible to ground-based personnel. Until recently, these estimates have been calculated without providing a measure of the variance or in large homogeneous forested areas where the variability would be minimal. This paper uses a Random Forest algorithm to produce estimates of QMDBH (cm) and volume (m3 ha-1) and a k-nn technique to produce variance estimates in a previously unproven heterogeneous forest environment. The area of interest (AOI) was the commercial forest in the Slieve Bloom Mountains in the Republic of Ireland, where the majority species are Sitka spruce (Picea sitchensis (Bong.) Carr.) and Lodgepole pine (Pinus contorta Dougl.). Field plots were measured during the Summer in 2018 during which a LiDAR campaign was flown and Sentinel 2 satellite imagery captured, which were both used as auxiliary variables. Root mean squared errors (rmse) for the modeled estimates of QMDBH and volume were 4.4 cm and 111 m3 ha-1 and the r2 values were 0.65 and 0.76 respectively. An independent dataset of pre-harvest forest stands was used to validate the modeled variance estimates. The results showed that 73% of QMDBH measurements and 60% of volume measurements were within the confidence intervals of the estimated values. The mean percentage standard deviation for QMDBH and volume were 17% and 21% respectively. This application of variance estimation in such a heterogeneous forest landscape adds further weight to the applicability of the methodology in a range of forest landscapes. This finding is extremely important as it highlights the benefits of using earth observation data to produce estimates and confidence intervals for small areas in heterogeneous forest landscape.