Early Monitoring Dataset for Pine Wilt Disease (PWD)
Description
This dataset contains close-range, time-series hyperspectral data, disease progression data, and high-definition RGB photographs acquired following the artificial inoculation of pine trees with the pine wood nematode. This study primarily utilized these data to conduct early dynamic monitoring of Pine Wilt Disease (PWD), to elucidate the dynamic patterns of its early development, and to determine the optimal timing for early detection.The dataset is composed of two main parts: code and data. The code section includes scripts for time-series sensitivity analysis, Mann-Kendall (M-K) trend and change-point detection, time-series analysis based on chromaticity indicators, and time-series correlation analysis between disease progression and typical spectral indicators. The data section primarily includes: the basic sample data used to calculate the Average Time-series Sensitivity Index (ATSI); the basic sample data for the M-K trend and change-point tests; on-site high-definition RGB photographs and their corresponding HSV space standard data; and the measured data of PWD infection progression.
Files
Steps to reproduce
The use of this dataset strictly follows the methodology detailed in the manuscript. The workflow is as follows: (1) Spectral Index Calculation: Based on the spectral sample data provided in the "Data" folder, the corresponding values for 20 spectral indices were calculated for each sample using their respective standard formulas. These calculated time-series values were then saved as separate data files, forming the basis for the subsequent sensitivity and trend analyses. (2) Time-Series Sensitivity Analysis: The Average Time-series Sensitivity Index (ATSI) was computed for all 20 spectral indices using the ATSI_cal script to quantify their effectiveness in tracking early-stage PWD. For each time-series, the data were first normalized to a [0, 1] range. Subsequently, the Normalized Standard Deviation (NSD), representing "overall fluctuation," and the Mean Absolute Successive Difference (MASD), representing "instantaneous sensitivity," were calculated. The ATSI for each index was then derived from these metrics, providing a quantitative measure of its overall temporal sensitivity. (3) M-K Trend and Change-Point Detection: The M-K trend and change-point tests were applied to the three most sensitive indices identified by the ATSI analysis (CI, WASCOSBNDI, and PRI) to determine their trend and the specific time of abrupt change. A significance level of α = 0.05 was used. The Z statistic was used to assess the significance of the trend direction. The change-point was identified as the first intersection of the forward (UFm) and backward (UBm) statistical sequences that fell within the critical lines (Y = ±1.96), pinpointing the onset of the significant change. (4) Chromaticity Indicator Analysis: The Chromaticity_Analysis script was used for RGB-to-HSV color space transformation, followed by Linear Discriminant Analysis (LDA) to measure the separability between sample groups. (5) Disease Progression and Correlation Analysis: After regression modeling of the disease progression data, the Correlation_Analysis script was used to calculate the correlation between the PRI and disease progression, linking the significant correlation timing to the optimal detection window. All data and code used for the analysis are publicly available via Mendelely Data to ensure reproducibility. The complete on-site high-definition RGB dataset (>20 GB) is available from the corresponding author upon reasonable request, as only a subset could be uploaded due to repository storage limitations.
Institutions
- University of the Chinese Academy of SciencesBeijing, Beijing