Data assimilation techniques were used to estimate forest stand data in 2011 by sequentially combining remote sensing based estimates of forest variables with predictions from growth models. Estimates of stand data, based on canopy height models obtained from image matching of digital aerial images at six different time-points between 2003 and 2011, served as input to the data assimilation. The assimilation routines were built on the extended Kalman filter. The study was conducted in hemi-boreal forest at the Remningstorp test site in southern Sweden (lat. 13°37′ N; long. 58°28′ E). The assimilation results were compared with two other methods used in practice for estimation of forest variables: the first was to use only the most recent estimate obtained from remotely sensed data (2011) and the second was to forecast the first estimate (2003) to the endpoint (2011). All three approaches were validated using nine 40 m radius validation plots, which were carefully measured in the field. The results showed that the data assimilation approach provided better results than the two alternative methods. Data assimilation of remote sensing time series has been used previously for calibrating forest ecosystem models, but, to our knowledge, this is the first study with real data where data assimilation has been used for estimating forest inventory data. The study constitutes a starting point for the development of a framework useful for sequentially utilizing all types of remote sensing data in order to provide precise and up-to-date estimates of forest stand parameters.

The use of multi-temporal laser scanner data is potentially an efficient method for monitoring of vegetation changes, for example, at the alpine treeline. Methods for relative calibration of multi-temporal airborne laser scanning (ALS) data sets and detection of experimental changes of tree cover in the forest–tundra ecotone was tested in northern Sweden (68° 20′ N, 19° 01′ E). Trees were either partly or totally removed on 6 m radius sample plots to simulate two classes of biomass change. Histogram matching was successfully used to calibrate the laser metrics from the two data sets and sample plots were then classified into three change classes. The proportion of vegetation returns from the canopy was the most important explanatory variable, which provided an overall accuracy of 88%. The classification accuracy was clearly dependent on the density of the forest.

Today, non-expensive remote sensing (RS) data from different sensors and platforms can be obtained at short intervals and be used for assessing several kinds of forest characteristics at the level of plots, stands and landscapes. Methods such as composite estimation and data assimilation can be used for combining the different sources of information to obtain up-to-date and precise estimates of the characteristics of interest. In composite estimation a standard procedure is to assign weights to the different individual estimates inversely proportional to their variance. However, in case the estimates are correlated, the correlations must be considered in assigning weights or otherwise a composite estimator may be inefficient and its variance be underestimated. In this study we assessed the correlation of plot level estimates of forest characteristics from different RS datasets, between assessments using the same type of sensor as well as across different sensors. The RS data evaluated were SPOT-5 multispectral data, 3D airborne laser scanning data, and TanDEM-X interferometric radar data. Studies were made for plot level mean diameter, mean height, and growing stock volume. All data were acquired from a test site dominated by coniferous forest in southern Sweden. We found that the correlation between plot level estimates based on the same type of RS data were positive and strong, whereas the correlations between estimates using different sources of RS data were not as strong, and weaker for mean height than for mean diameter and volume. The implications of such correlations in composite estimation are demonstrated and it is discussed how correlations may affect results from data assimilation procedures.

Airborne Laser Scanning (ALS) has implied a disruptive transformation of how data are gathered for forest management planning in Nordic countries. We show in this study that the accuracy of ALS predictions of growing stock volume can be maintained and even improved over time if they are forecasted and assimilated with more frequent but less accurate remote sensing data sources like satellite images, digital photogrammetry, and InSAR. We obtained these results by introducing important methodological adaptations to data assimilation compared to previous forestry studies in Sweden. On a test site in the southwest of Sweden (58°27′N, 13°39′E), we evaluated the performance of the extended Kalman filter and a proposed modified filter that accounts for error correlations. We also applied classical calibration to the remote sensing predictions. We evaluated the developed methods using a dataset with nine different acquisitions of remotely sensed data from a mix of sensors over four years, starting and ending with ALS-based predictions of growing stock volume. The results showed that the modified filter and the calibrated predictions performed better than the standard extended Kalman filter and that at the endpoint the prediction based on data assimilation implied an improved accuracy (25.0% RMSE), compared to a new ALS-based prediction (27.5% RMSE).

The statistical framework of data assimilation provides methods for utilizing new data for obtaining up-to-date forest data: existing forest data are forecasted and combined with each new remote sensing data set. This new paradigm for updating forest database, well known from other fields of study, will provide a framework for utilizing all available remote sensing data in proportion to their quality to improve prediction. It also solves the problem that not all remote sensing data sets provide information for the entire area of interest, since areas with no remote sensing data can be forecasted until new remote sensing data become available. In this study, extended Kalman filtering was used for assimilating data from 19 TanDEM-X InSAR images on 137 sample plots, each of 10-meter radius at a test site in southern Sweden over a period of 4 years. At almost all time points data assimilation resulted in predictions closer to the reference value than predictions based on data from that single time point. For the study variables Lorey's mean height, basal area, and stem volume, the median reduction in root mean square error was 0.4 m, 0.9 m²/ha, and 15.3 m³/ha (2, 3, and 6 percentage points), respectively.