Forest Inventory Using Radar and Spectral Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current forest inventory methods are costly and inefficient, particularly when aiming for high accuracy in assessing the number, size, and species of trees, as they often require extensive ground-based surveys or expensive lidar technology, and existing remote sensing methods struggle with direct correlations between radar data and basal area measurements.
Innovation Solution
A computer-implemented method using a combination of radar data, spectral images, and minimal ground-based survey data, where each pixel in a spectral image is treated as a potential stand, with coefficients estimated to construct tree parameters, reducing the need for individual tree measurements and minimizing ground measures by matching sample plots to pixel data distributions, thereby improving estimation accuracy at a larger scale.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ground-based surveys are used to extensively survey the entire plot, then measurement precision is improved, but cost increases significantly
Solution Approach 1:
The patent uses remote sensing data (satellite images, aerial photography, radar data) to create a copy or representation of the forest plot, allowing inventory assessment without physically visiting every location. This copying approach maintains measurement precision while dramatically reducing the cost and effort required compared to exhaustive ground-based surveys.
Solution Approach 2:
The patent introduces remote sensing data as an intermediary between the forest plot and the inventory assessment process. Instead of direct ground-based measurement of every tree, the system uses intermediate data layers (spectral images, radar data, digital elevation models) that can be processed to derive forest characteristics, thereby reducing cost while maintaining accuracy.
2Measurement precision
If lidar technology is used for high accuracy assessment, then measurement precision is improved, but cost increases
Solution Approach 1:
The patent merges multiple types of remote sensing data (optical satellite images, radar data, digital elevation models) with limited ground-based survey data to create a comprehensive forest inventory system. This combination allows the system to achieve accuracy comparable to lidar technology without incurring lidar costs, as the synergistic integration of multiple data sources compensates for the absence of expensive lidar measurements.
Solution Approach 2:
The patent creates a multi-functional system that can assess various forest parameters (tree species, diameter, height, volume, biomass) using a single integrated approach combining remote sensing and statistical methods. This universal system replaces the need for multiple specialized technologies including expensive lidar, providing cost-effective multi-parameter assessment.
3Quantity of substance
If remote sensing methods are used, then cost is reduced, but measurement precision deteriorates due to lack of direct correlation with basal area
Solution Approach 1:
The patent uses spectral indices and radar backscatter coefficients as intermediary variables that indirectly correlate with basal area. Instead of relying on direct correlation between remote sensing data and basal area, the system introduces intermediate statistical models and proxy measurements that bridge the gap, allowing accurate basal area estimation from remote sensing data while maintaining cost efficiency.
Solution Approach 2:
The patent transforms remote sensing data parameters (spectral reflectance, radar backscatter) into forest inventory parameters (basal area, volume, biomass) through statistical relationships and calibration with ground data. By changing and reinterpreting the parameters through multiple transformation steps, the system overcomes the lack of direct correlation and achieves precise basal area estimation.
4Measurement precision
If sampling intensity is increased to improve quality, then measurement precision is improved, but cost increases
Solution Approach 1:
The patent uses remote sensing data to create comprehensive copies of the entire study area, allowing assessment of all pixels rather than relying on limited ground samples. This copying approach enables high confidence levels for the entire area without the prohibitive cost of intensive sampling, as the remote sensing data provides area-wide coverage that substitutes for multiple ground samples.
Solution Approach 2:
The patent transitions from two-dimensional ground-based sampling to three-dimensional assessment by incorporating vertical information from radar data and digital elevation models. This dimensional enhancement allows the system to derive accurate forest parameters from remote sensing data without increasing ground sampling intensity, as the additional dimensional information compensates for reduced sampling effort.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
Methods and systems are provided that inventory a plot of trees based on data including one or more (e.g., all) of radar images of the plot, spectral images of the plot (e.g., high resolution images taken by satellite), other data (e.g., elevation, slope, aspect), and actual tree survey data physically collected about the plot and/or another plot having similar characteristics. Although the actual tree survey data collected is typically less than the amount of actual survey data used by prior approaches, the present systems and methods are still capable of inventorying the entire plot with a high degree of confidence (e.g., at least 95% confidence).