Name
Title | ||
|---|---|---|
Forest Aboveground Biomass Estimation Using Multi-Features Extracted By Fitting Vertical Backscattered Power Profile Of Tomographic Sar |
Abstract | ||
|---|---|---|
Forest aboveground biomass (AGB), which plays an important role in the study of global carbon cycle, is one of the most important indicators in forest resource monitoring. Thus, how to estimate and map regional forest AGB quickly and accurately attracts more interests of researchers. Tomographic SAR (TomoSAR) is an advanced SAR technique developed in recent years, which has a wide range application in forest AGB estimation. In this paper, we proposed a multi-feature-based modeling method to estimate forest AGB by fitting backscattered power of TomoSAR vertical profile. The procedure of the proposed method includes four parts: (1) Processing TomoSAR data to obtain the backscattered power of vertical profile. (2) Fitting the backscattered power of the vertical profile. (3) Analyzing the fitted backscattered power distribution characteristic of the vertical profile. (4) Extracting the TomoSAR vertical profile features according to the forest AGB measurement factors based on the dendrometry theory. In this paper, we proposed two new features like the forest average height weighted by backscattered power (BPFAH) and the total length of the backscattered power curve (LBPC) as supplement features to estimate forest AGB by TomoSAR technique. We also used the traditional TomoSAR features including backscattered power at specific height layer of vertical power profile (BPV) and forest average height (FAH) for AGB estimation. After the feature selection, the selected features and the ground field data of the forest AGB were used for regression and modeling. Then the forest AGB was estimated and the accuracy was validated. The results showed that the accuracy of proposed method is 90.73%, and RMSE is 42.45 t/ha. Finally, we discussed the performance of our proposed method compared with traditional methods. |
Year | DOI | Venue |
|---|---|---|
2021 | 10.3390/rs13020186 | REMOTE SENSING |
Keywords | DocType | Volume |
TomoSAR, forest AGB, backscattered power, forest vertical profile | Journal | 13 |
Issue | Citations | PageRank |
2 | 0 | 0.34 |
References | Authors | |
0 | 6 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Xiangxing Wan | 1 | 0 | 0.34 |
| Zengyuan Li | 2 | 0 | 1.35 |
| Erxue Chen | 3 | 2 | 3.45 |
| Lei Zhao | 4 | 0 | 1.01 |
| Wangfei Zhang | 5 | 1 | 1.39 |
| Kunpeng Xu | 6 | 0 | 0.68 |