A MWCMLAI-Net method for LAI inversion in maize and rice using GF-3 and Lutan radar data

Xiaoxuan Wang,Xiaoping Lu,Zenan Yang
DOI: https://doi.org/10.1080/17538947.2024.2341128
IF: 4.606
2024-04-13
International Journal of Digital Earth
Abstract:This study aimed at alleviating the problems of unsatisfactory inversion accuracy and weak model stability in LAI remote sensing quantitative inversion. The properties and complex scattering mechanism of SAR data specify the polarization combinations and frequencies. This paper proposes an improved water cloud model combined with a deep neural network (MWCMLAI-Net) for high-precision inversion. The polarized GF-3 (C-band) and Lutan (L-band) were used to investigate the potential of SAR images to estimate LAI, a strong indicator of crop productivity. The study selected xiangfu district in the eastern part of Kaifeng City, Henan Province, as the test area and investigated the LAI of maize and rice. The RVIFreeman Model, backward scattering coefficient extracted by the modified cloud and water model (MWCM), and LAI obtained by the inversion of the MWCM were used as the inputs, and the MWCMLAI-Net inversion of the LAI was constructed. The results showed that the model's inverted LAI fitting accuracies of maize and rice for the three fertility periods were better than the other models, with R 2 above 0.8516 and RMSE below 0.3999 m 2 /m 2 . The addition of noise did not affect the results.
geography, physical,remote sensing
What problem does this paper attempt to address?
### Problems the Paper Attempts to Solve This paper aims to address the issues of low accuracy and poor model stability in the current quantitative remote sensing inversion of Leaf Area Index (LAI). Specifically, the paper proposes an improved water cloud model combined with a deep neural network (MWCMLAI-Net) method, utilizing Gaofen-3 (GF-3) radar and Lutan radar data to enhance the accuracy of LAI inversion. #### Research Background - **Importance of LAI**: LAI is a crucial parameter for describing vegetation growth and plays a significant role in the carbon and water cycles. It is also an important indicator for measuring crop yield, photosynthetic capacity, and growth health. - **Limitations of Traditional Methods**: Traditional ground measurement methods can obtain LAI values for local areas but are time-consuming and difficult to apply on a large scale. Optical remote sensing images have low inversion accuracy and poor reliability under cloud cover conditions. #### Research Objectives - **Improve Inversion Accuracy**: By combining multi-polarization GF-3 (C-band) and HH polarization Lutan (L-band) radar data, the study aims to improve the inversion accuracy of LAI for corn and rice. - **Enhance Model Stability**: The proposed MWCMLAI-Net model not only improves inversion accuracy but also has strong noise resistance and stability. #### Method Overview - **Experimental Area**: The study selected Xiangfu District in the eastern part of Kaifeng City, Henan Province, as the experimental area and chose three growth stages for LAI inversion analysis of corn and rice. - **Data Processing**: Preprocessing steps such as radiometric correction and geometric correction were performed on GF-3 and Lutan radar data. - **Model Construction**: The improved water cloud model was used to extract backscatter coefficients, which were then combined with LAI and input into the MWCMLAI-Net model for nonlinear mapping. #### Results Validation - **Results Evaluation**: The model's LAI inversion results for corn and rice in the three growth stages outperformed other models, with correlation coefficients (R²) all above 0.8516 and root mean square errors (RMSE) below 0.3999 m²/m². Through these methods, the paper aims to provide an efficient and stable LAI inversion method to meet the needs of agricultural monitoring and biogeochemical cycle research.