Crop loss identification at field parcel scale using satellite remote sensing and machine learning.

Identifying crop loss at field parcel scale using satellite images is challenging: first, crop loss is caused by many factors during the growing season; second, reliable reference data about crop loss are lacking; third, there are many ways to define crop loss. This study investigates the feasibility of using satellite images to train machine learning (ML) models to classify agricultural field parcels into those with and without crop loss. The reference data for this study was provided by Finnish Food Authority (FFA) containing crop loss information of approximately 1.4 million field parcels in Finland covering about 3.5 million ha from 2000 to 2015. This reference data was combined with Normalised Difference Vegetation Index (NDVI) derived from Landsat 7 images, in which more than 80% of the possible data are missing. Despite the hard problem with extremely noisy data, among the four ML models we tested, random forest (with mean imputation and missing value indicators) achieved the average AUC (area under the ROC curve) of 0.688±0.059 over all 16 years with the range [0.602, 0.795] in identifying new crop-loss fields based on reference fields of the same year. To our knowledge, this is one of the first large scale benchmark study of using machine learning for crop loss classification at field parcel scale. The classification setting and trained models have numerous potential applications, for example, allowing government agencies or insurance companies to verify crop-loss claims by farmers and realise efficient agricultural monitoring.

A Clustering Framework for Monitoring Circadian Rhythm in Structural Dynamics in Plants From Terrestrial Laser Scanning Time Series.

Terrestrial Laser Scanning (TLS) can be used to monitor plant dynamics with a frequency of several times per hour and with sub-centimeter accuracy, regardless of external lighting conditions. TLS point cloud time series measured at short intervals produce large quantities of data requiring fast processing techniques. These must be robust to the noise inherent in point clouds. This study presents a general framework for monitoring circadian rhythm in plant movements from TLS time series. Framework performance was evaluated using TLS time series collected from two Norway maples (Acer platanoides) and a control target, a lamppost. The results showed that the processing framework presented can capture a plant's circadian rhythm in crown and branches down to a spatial resolution of 1 cm. The largest movements in both Norway maples were observed before sunrise and at their crowns' outer edges. The individual cluster movements were up to 0.17 m (99th percentile) for the taller Norway maple and up to 0.11 m (99th percentile) for the smaller tree from their initial positions before sunset.

Pre-crop Values From Satellite Images for Various Previous and Subsequent Crop Combinations.

Monocultural land use challenges sustainability of agriculture. Pre-crop value indicates the benefits of a previous crop for a subsequent crop in crop sequencing and facilitates diversification of agricultural systems. Traditional field experiments are resource intensive and evaluate pre-crop values only for a limited number of previous and subsequent crops. We developed a dynamic method based on Sentinel-2 derived Normalized Difference Vegetation Index (NDVI) values to estimate pre-crop values on a field parcel scale. The NDVI-values were compared to the region specific 90th percentile of each crop and year and thereby, an NDVI-gap was determined. The NDVI-gaps for each subsequent crop in the case of monocultural crop sequencing were compared to that for other previous crops in rotation and thereby, pre-crop values for a high number of previous and subsequent crop combinations were estimated. The pre-crop values ranged from +16% to -16%. Especially grain legumes and rapeseed were valuable as pre-crops, which is well in line with results from field experiments. Such data on pre-crop values can be updated and expanded every year. For the first time, a high number of previous and following crop combinations, originating from farmer's fields, is available to support diversification of currently monocultural crop sequencing patterns in agriculture.