< Back to previous page

Project

High spatial resolution satellite imagery for irrigation scheduling in hedgerow cropping systems

Over the last decades, the agricultural sector has shifted to produce more with less inputs. This has led to the development of precision horticulture, an information-based farm management concept that requires frequent monitoring of biophysical, structural and environmental variables at high spatial and temporal scales. In situ monitoring is however time consuming and labor intensive, hampering the number of samples and repetitions. An alternative can be found in remote sensing technology. In this dissertation, the application of remote sensing for irrigation scheduling in commercial pear orchards was investigated. More specifically, the research focused on the possibilities of high spatial resolution satellite sensors for regulated deficit irrigation. Problems and bottlenecks related to the practical application of high spatial resolution satellite imagery were highlighted and tackled.
Remote sensing as a proxy for in situ measurements should provide accurate and useful information for irrigation scheduling purposes and ideally be based on reliable substitutes for primary indicators of water stress. The lack of robust spectral indicators of water status within the visible and near infrared region of the electromagnetic radiation was addressed. The red-edge region was found to show significant correlations with in situ measured indicators and provided a good estimation of the overall health and stress level of pear canopies.
Satellite derived remote sensing imagery over orchards contained mixtures of canopies and backgrounds. The discontinuous canopy cover fraction for these heterogeneous cropping systems as a result of age, growing system, tree and row spacing caused significant differences to remote sensing products within and between orchards. The lack of a generic solution was addressed and a novel vegetation index correction method was presented to resolve the canopy fraction distribution for high spatial resolution satellite imagery over hedgerow orchards. Variations caused by canopy cover fractions, age and growing system were removed.
Varying viewing angles within and between satellite imagery time series obstructed the relationship between biophysical variables and spectral measurements. The view-angle sensitivity of common vegetation indices for high spatial resolution imagery of hedgerow cropping systems was quantified in relation to the estimation of biophysical variables. The results have shown the necessity of vegetation index selection for variable viewing applications to obtain an optimal derivation of biophysical variables in all circumstances.
The influence of contributing factors on fruit yield and quality varied during different phenological stages. As a result, the relationship between spectral measurements and production variables was linked to different phenological stages. This would require the use of different vegetation indices during different growing stages. To optimize remote sensing monitoring throughout the growing season, the temporal profile of the association between spectral information and production variables was quantified. Results indicated that optimal time frames for remote yield prediction in irrigated orchards would fall outside strong vegetative growth periods, while measurements at the end of fruit fall will result in optimal estimations for rainfed orchards.
In the end, some of the stumbling blocks and influences were addressed and made it possible to distinguish and quantify the temporal influence of vegetation indices towards the estimation of water status and the prediction of crop yield. These understandings should pave the way for an eventual implementation of data driven decision support systems within the agricultural production system and in the end producing more crop per drop. However, prior to practical application, future work should be focused on improving methodology and processing, the integration with other sources and availability.

Date:1 Oct 2010 →  3 Oct 2018
Keywords:Satellite imagery, Irrigation, Orchard
Disciplines:Control systems, robotics and automation, Design theories and methods, Mechatronics and robotics, Computer theory, Agriculture, land and farm management, Biotechnology for agriculture, forestry, fisheries and allied sciences, Fisheries sciences
Project type:PhD project