Data underlying the VIOLA model for potato stress classification using NDVI

This data were generated in the context of the Kennis Basis 38 Program on Data Driven and High Tech - use case on AI on arable and animal farms.The objective of the experiment was to evaluate the effect of various potato stress factors on canopy reflectance.This data have been collected to validate a model trained on a synthetic dataset.The experiment has been conducted for two growing seasons: 2023 and 2024.  The experiment was a factorial combination of 2 cultivars (Frieslander early cultivar and Avenger late cultivar) x 5 treatments. In 2023 we had no replicates whereas in 2024 we had 3 replicates.<br>The treatments applied to this experiment were:* No stress* Nitrogen stress* Poor potato emergence* Phytophtora (only in 2024)* Weeds                                                                                                                     The original experiment was supposed to inclide also a water stress treatment by irrigating all fields except for the water stress treatment, however 2023 and 2024 were two wet years and thus irrigation was not needed, this implies that we have two controls treatments instead of one. We measured reflectance using hyperspectral point sensor (TEC5) and drone. For the drone here we report the average reflectance by plot, date and band, drone images can be made available upon request.

本数据集依托面向耕地与畜牧农场人工智能应用的案例研究，由Kennis Basis 38数据驱动与高科技项目（Kennis Basis 38 Program on Data Driven and High Tech）资助生成。本实验旨在评估多种马铃薯胁迫因子对冠层反射率的影响，所获数据用于验证基于合成数据集训练得到的模型。

实验覆盖2023与2024两个生长季，采用双因素析因设计：2个马铃薯品种（早熟品种Frieslander与晚熟品种Avenger）×5种胁迫处理。其中2023年未设置生物学重复，2024年设置3次重复。

本次实验设置的处理如下：
- 无胁迫
- 氮素胁迫
- 马铃薯出苗不齐
- 疫霉菌胁迫（Phytophtora，仅2024年设置）
- 杂草胁迫

原实验计划增设水分胁迫处理组，即除水分胁迫组外其余地块均实施灌溉，但2023与2024年均为多雨年份，无需额外灌溉，因此实际最终设置了两组对照处理而非一组。

本研究采用高光谱点传感器（TEC5）与无人机开展反射率数据采集。本次报告的无人机数据为各样地、各观测日期对应波段的平均反射率，无人机影像可应研究需求提供。