Partner und Internationale Organisationen
(Englisch)
|
AT, CH, CZ, DE, DK, EE, EL, ES, FI, FR, HU, IE, IS, IT, LV, NL, NO, PL, PT, RO, RS, SE, SI, UK
|
Abstract
(Englisch)
|
The goal of this project was to develop and test estimation procedures for the Swiss National Forest Inventory (NFI) under the new annual design with a focus on preserving desireable design-based properties of the estimators. Under the annual design one ninth of the permanent terrestrial plots are sampled every year rather than all together every 10 years. The main advantage of this is that estimates can be obtained continuously rather than periodically. The disadvantage is that with a time point grouping of 3 years, the variance increases by approximately 3 times. This project explored ways to incorporate remote sensing data, previous and updated plot measurements to lessen this increase in variance with two- and three-phase regression estimators. Three-phase regression estimators were specifically tested for stem volume and biomass of standing living trees (with DBH threshold of 12cm) using data from the 2nd and 3rd Swiss national forest inventories. The first phase was comprised of manually interpreted aerial photographs with state and change of canopy height information. The second phase consisted of the previous measurements of all terrestrial plots, and the third phase was the ground truth available for 3 years of annually available data. Two-phase regression was used to assess the contribution of the first and second phases of auxiliary data to variance reduction. It was demonstrated that the increase in variance due to the transition to the annual design can be roughly cut in half using a three-phase regression estimators for volume and biomass estimation. The previous measurement was a moderately stronger predictor than remote sensing data alone, and the estimates can be improved with the change in crown cover measured on aerial photographs and terrestrial plots updated with single tree growth models. Methods and detailed results are summarized in an article published in the Canadian Journal of Forest Research entitled, “Integrating remote sensing and past inventory data under the new annual design of the Swiss National Forest Inventory using three-phase design-based regression estimation.'
|
