Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #377832

Research Project: Integrating Remote Sensing, Measurements and Modeling for Multi-Scale Assessment of Water Availability, Use, and Quality in Agroecosystems

Location: Hydrology and Remote Sensing Laboratory

Title: Time-series clustering of remote-sensing retrievals for defining management zones in a vineyard

Author
item OHANA-LEVI, N. - Ben Gurion University Of Negev
item Gao, Feng
item Knipper, Kyle
item Kustas, William - Bill
item Anderson, Martha
item ALSINA, M. - E & J Gallo Winery
item SANCHEZ, L. - E & J Gallo Winery
item KAMELI, A. - Ben Gurion University Of Negev

Submitted to: Remote Sensing of Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/5/2021
Publication Date: 10/18/2021
Citation: Ohana-Levi, N., Gao, F.N., Knipper, K.R., Kustas, W.P., Anderson, M.C., Alsina, M., Sanchez, L., Kameli, A. 2021. Time-series clustering of remote-sensing retrievals for defining management zones in a vineyard. Remote Sensing of Environment. https://doi.org/10.1007/s00271-021-00752-0.
DOI: https://doi.org/10.1007/s00271-021-00752-0

Interpretive Summary: A standard management technique for applying precision agriculture is to partition an agricultural field into multiple homogenous management zones (MZs), which enable various site-specific management applications, such as irrigation, fertilization, and pest control. Usually, MZs are determined using static field attributes, such as terrain, soil, and plant characteristics. Seasonal patterns of crop conditions have not been considered in traditional approaches. This study proposed an approach for generating MZs using a remote-sensing timeseries clustering technique. The approach was applied to a California vineyard during four growing seasons (2015-2018) using evapotranspiration (ET), leaf area index (LAI), and normalized difference vegetation index (NDVI) time-series images. Results show that remote-sensing retrievals of ET, LAI, and NDVI were suitable choices for partitioning the vineyard into MZs. The proposed approach is highly effective when the delineation of the field is conducted based on the seasonal patterns of crop conditions. The proposed framework is applicable for other cases in both agricultural systems and environmental modeling.

Technical Abstract: Management zones (MZs) are efficient for applying site-specific management (SSM) in agricultural fields. This study proposes an approach for generating MZs using time-series clustering (TSC) to enable SSM and time-specific management (TSM). TSC was applied to time series of remote sensing images in a California vineyard during four growing seasons (2015-2018) using three datasets, including evapotranspiration (ET), leaf area index (LAI), and normalized difference vegetation index (NDVI). The MZs maps and their temporal dynamics were compared. A dissimilarity index helped determine the optimal number of clusters and to compare the TSC results. The differences between the cluster centers of the different clusters (MZs) were calculated, along with the ratio between the centers’ difference and the range of each dataset. The cluster centers and their differences were normalized and similarities were compared using Fréchet distance. The MZs were compared using Cramer’s V. An aggregated MZs map was generated using multivariate clustering. The findings show that the LAI TSC outperformed the other datasets, with a stronger cluster separation. The MZs maps of NDVI and LAI were nearly identical (Cramer’s V of 0.97), while ET showed weaker similarities to NDVI and LAI (0.61 and 0.62, respectively). Similar findings were observed for the Fréchet distances. The aggregated MZs map was composed of ET and NDVI, with spatial patterns similar to a 2016 yield map. With increasing availability of higher spatial and temporal resolution satellite imagery, TSC may be further utilized for defining within-field spatial variability and temporal dynamics for SSM and TSM.