研究目的
To identify spectral areas sensitive to rice leaf folder damage at the jointing stage and evaluate the use of vegetation indices for monitoring damage at the full heading stage using GF-1 WFV data.
研究成果
Hyperspectral data and GF-1 satellite images are feasible for monitoring and warning rice leaf folder outbreaks. SAVI showed the highest correlation with pest population, indicating potential for dynamic damage monitoring, though ground survey data are still needed for large-area models.
研究不足
Uncertainties in estimation accuracy due to factors like drought and waterlogging that can affect vegetation indices. Future studies should explore other remote sensing data (e.g., SAR, other optical satellites) to exclude confounding factors.
1：Experimental Design and Method Selection:
The study used remote sensing methods, including hyperspectral data collection with an ASD spectroradiometer and satellite imagery analysis with GF-1 data, to monitor rice leaf folder damage. Vegetation indices (RVI, NDVI, EVI, EVI2, SAVI, OSAVI) were calculated and correlated with pest populations.
2：Sample Selection and Data Sources:
Nine rice fields in Xinghua City, Jiangsu Province, were selected based on damage severity levels at the full heading stage in 2015. Pest population data were collected through field surveys by professionals.
3：Pest population data were collected through field surveys by professionals.
List of Experimental Equipment and Materials:
3. List of Experimental Equipment and Materials: ASD FieldSpec Hand Held Spectroradiometer, white BaSO4 calibration panel, GF-1 satellite images, and software (ViewSpecPro, ENVI) for data analysis.
4：Experimental Procedures and Operational Workflow:
Canopy spectra were measured at 12:00–14:00 under cloudless and windless conditions using the spectroradiometer placed 0.6 m above the canopy. Ten measurements per plot were averaged. GF-1 images were processed through radiometric calibration, atmospheric correction, and geometric correction. Vegetation indices were derived and correlated with pest populations using Pearson correlation.
5：6 m above the canopy. Ten measurements per plot were averaged. GF-1 images were processed through radiometric calibration, atmospheric correction, and geometric correction. Vegetation indices were derived and correlated with pest populations using Pearson correlation.
Data Analysis Methods:
5. Data Analysis Methods: Reflectance data from 400-1000 nm were analyzed with ViewSpecPro. Correlation analyses were performed using Pearson method at 99% significance level. Vegetation indices were calculated from GF-1 data and correlated with pest populations.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
