Skip to main content
ARS Home » Southeast Area » Jonesboro, Arkansas » Delta Water Management Research » Research » Publications at this Location » Publication #419876
Research Project: Optimizing the Management of Irrigated Cropping Systems in the Lower Mississippi River Basin

Location: Delta Water Management Research

Title: The first fine-resolution mapping of contour-levee irrigation using deep Bi-Stream convolutional neural networks

Author
item LIANG, LU - University Of North Texas
item MEYARIAN, ABOLFAZL - University Of North Texas
item YUAN, XIAOHUI - University Of North Texas
item RUNKLE, BENJAMIN - University Of Arkansas
item MIHAILA, GEORGE - University Of North Texas
item QIN, YUCHU - Institute Of Remote Sensing And Digital Earth, Chinese Academy Of Sciences
item DANIELS, JACOB - University Of North Texas
item Reba, Michele
item Rigby Jr, James

Submitted to: International Journal of Applied Earth Observation and Geoinformation
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/16/2021
Publication Date: 11/23/2021
Citation: Liang, L., Meyarian, A., Yuan, X., Runkle, B.R., Mihaila, G., Qin, Y., Daniels, J., Reba, M.L., Rigby Jr, J.R. 2021. The first fine-resolution mapping of contour-levee irrigation using deep Bi-Stream convolutional neural networks. International Journal of Applied Earth Observation and Geoinformation. 105(102631). https://doi.org/10.1016/j.jag.2021.102631.
DOI: https://doi.org/10.1016/j.jag.2021.102631

Interpretive Summary: The production of rice requires more water than other common crops in the Mid-South United States. Water use in rice production is influenced by field slope, water application method, and how water is managed. More than half of the rice fields in Arkansas, the largest rice growing state in the United States, use contour-levee cascade irrigation. Determining the distribution and adoption of this important irrigation method is difficult to do with remote sensing. A new remote sensing method that uses a bi-stream encoder-decoder architecture to capture spectral information and gradient features was employed. Images were tested across 10 Arkansas counties with average accuracy of 86% and the new method made improved over benchmark methods. Identifying where contour-levee cascade irrigation fields are located help inform land managers of water needs and improve planning of valuable water resources.

Technical Abstract: Agricultural irrigation accounts for nearly 70% of global freshwater withdrawal. Among irrigation practices, contour-levee cascade irrigation is of particular interest as it is water-intensive and widely used in many rice production regions. Despite its significant environmental implications, no study has quantified the distribution of contour-levee irrigation. One major challenge of remote sensing-based contour-levee field detection is how to accurately identify the thin and curved levee lines whose appearance varies dramatically in different fields. This paper presents a new deep network-based method that jointly optimizes semantically meaningful features to quantify the contour-levee fields. This new method uses a bi-stream encoder-decoder architecture to capture spectral information and gradient features. To maintain image gradient sharpness, a skip connection approach is employed to facilitate gradient propagation across long-range connections. Moreover, the new method uses deep supervision to generate more informative features from the earlier hidden layers and superpixel segmentation to reduce classification noise as a post-processing step. By testing against 41 images across 10 Arkansas counties, the average accuracy was 86.23% and the method achieved 15%-17% improvement over benchmark methods. The results show that IrrNet-Bi-Seg maintains good transferability and is thus promising for larger-scale applications.