| Accurate inversion of forest aboveground biomass are critical for many applications, including wildlife management and biodiversity studies, fire modeling, and carbon stock estimation. However, because of the lack of forest vegetation survey data, the sampling of forest vegetation aboveground biomass is devastating, and the biomass estimation technology is not mature enough, the accurate inversion of forest aboveground biomass remains a great challenge.In this paper, we choose the high resolution multispectral data and synthetic aperture radar data as the main data source, selected 50 samples of 30 × 30 m2 from subtropical forest fixed sample plot in Heishiding randomly, we calculate biomass of all the individual trees according to the specific allometric growth equation, extract the NDVI statistical indicators(maximum, minimum, mean and standard deviation) and PC1 statistical indicators(maximum, minimum, mean and standard deviation)of Worldview2 data in corresponding samples, and extract the ? statistical indicators(maximum, minimum, mean and standard deviation) of ALOS PALSAR data in corresponding samples, then we using the AIC rule to build inversion model of Worldview2 data and ALOS PALSAR data. We will apply the best aboveground biomass inversion model in the Heishiding nature reserve, to get the map of aboveground biomass distribution in the Heishiding nature reserve, then we analysis the influence of terrain factors on the aboveground biomass spatial distribution pattern in the Heishiding nature reserve. Main results are as follows:1. Based on the ALOS PALSAR data to construct the aboveground biomass inversion model, through the AIC rule to select the optimal parameters of ? statistics indicators, then we using the mean of ?, the standard deviation of ? to build aboveground biomass inversion model, the precision of the model validation coefficient reached 0.7361, RMSE is 52.1385 t/ha, overall degree of accuracy is 73.40%.2. Based on the Worldview2 data to construct the aboveground biomass inversion model, through the AIC rule to select the optimal parameters of NDVI and PC1 statistics indicators, then we using the mean of NDVI, the mean of PC1 to build aboveground biomass inversion model, the precision of the model validation coefficient reached 0.8086, RMSE is 38.4151 t/ha, overall degree of accuracy is 86.02%.3. Based on the Worldview2 data and ALOS PALSAR data to construct the aboveground biomass inversion model, through the AIC rule to select the optimal parameters of NDVI, PC1 and ? statistics indicators, then we using the mean of NDVI, the max of NDVI, the min of NDVI and the min of ? to build aboveground biomass inversion model, the precision of the model validation coefficient reached 0.8940, RMSE is 29.7497 t/ha, overall degree of accuracy is 90.62%. Studies have shown that, the correlation coefficient of multi-source data synergy to get the inversion model of aboveground biomass were higher than single data to get the inversion model of aboveground biomass, and the RMSE of multi-source data synergy to get the inversion model of aboveground biomass were lower than single data to get the inversion model of aboveground biomass, which improves the forest aboveground biomass inversion accuracy, contributes to the realization of accurate estimating and monitoring of forest aboveground biomass.4. The aboveground biomass distribution increase gradually from north to south in the Heishiding nature reserve, according to the division of nature reserve, from north to south, respectively is the buffer area, experimental area and the core area, in core area complete broad-leaved forest occupied about 40% of the total area in Heishiding nature reserve, the biomass of the broad-leaved forest area is the highest in the core area. The forest aboveground biomass distribution mainly concentrates within the range of 200-300- t/ha in the Heishiding nature reserve. The overall forest aboveground biomass predictive value is 253.33 t/ha in the Heishiding nature reserve. Terrain factors have significant correlation with the aboveground biomass distribution, the influence of terrain factors on the aboveground biomass distribution is as follows:(1) with the increase of altitude, the average aboveground biomass declines trend after the first increase in the research area, at the altitude of 700-750 m segment reach the maximum biomass of 295.84 t/ha;(2) As the change of slope, when the slope is less than 25 °, average aboveground biomass increased gradually with the increase of the slope, after the slope is greater than 25 °, average aboveground biomass declined gradually with the increase of the slope;(3) As the change of aspect, aboveground biomass distribution in each aspect upward followed by sunny, semi-sunny, semi-shady, shady, the average aboveground biomass of each aspect, in turn, is 259.71 t/ha, 216.27 t/ha, 158.22 t/ha and 144.88 t/ha. Average biomass research area is to present a clear distribution within the aspect, with the increase of solar radiation angle degree, the average aboveground biomass increased after the first decreases, reached the maximum of 269.40 t/ha in the south aspect. As the change of solar radiation angle degree, gradually turn to north aspect, the intensity of solar radiation is reduced, the vegetation photosynthesis of this region also weaken, average aboveground biomass decreased obviously. |
PDF Full Text Request