Development of a non-destructive ground-based nitrogen sensing system for coffee plantations | |
| Author | Putra, Bayu Taruna Widjaja |
| Call Number | AIT Diss. no.AE-17-02 |
| Subject(s) | Coffee plantations Nitrogen as fertilizer Soils--Nitrogen content |
| Note | A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Agricultural Systems and Engineering |
| Publisher | Asian Institute of Technology |
| Abstract | Coffee is an economically-important commodity for Indonesia; but 96 % of total coffee plantation areas are managed by smallholders. In the recent years, the Robusta coffee production has not been optimum compared with other coffee producing countries. One of the problems is inability of smallholders to access and afford modern technologies. Proper nutrient management and technological improvements in fertilizing could restore and accelerate the productivity of smallholders. In addition, proper use of fertilizer will not only improve soil nutrient levels and productivity of coffee plants, but will also control the pests and diseases attack. To bridge the technology gap for coffee farmers, we explored and developed low-cost tool using ground-based remote sensing for smallholders. In this study, we focused on Nitrogen (N), which is the most used nutrient by plants. N content in plant is an important factor that characterizes its yields and its health. Determining the status of N critical level often requires field observation. Although it can be done using conventional methods like destructive sampling, this method is still costly and time consuming. Other technologies like remote sensing applications (such as satellite, airborne and UAV) are still expensive for smallholders in developing country like Indonesia. Unlike cereal crops such as paddy, wheat and corn, the monitoring using remote sensing technology in Robusta coffee requires additional attention while measuring physical properties using non destructive measurement. These technologies have challenges due to narrow plantation area, hilly terrain, and covered by shade trees. The objective of this study was to develop a low-cost ground-based nitrogen sensing system for coffee plantation using non-destructive method. The use of consumer-grade cameras are explored as low-cost device for measuring nitrogen content of Robusta coffee plants. To achieve this objective, the consumer-grade digital cameras (unmodified and modified with different low-cost external filters like NIR-Red (R) and NIR-Red-edge (RE) filters) are proposed for direct-leaves and above-canopy measurements, and compared with different commercial sensors like SPAD-502 Chlorophyll meter, CropSpec® active canopy sensor, and Spectrometer to determine how well N critical level of Robusta coffee plants can be estimated with these non-destructive methods. The destructive method of plant tissue (leaves) chemical analysis is used to determine the actual value of nitrogen critical level. A series of experiments was conducted at Indonesian Coffee and Cocoa Research Institute (ICCRI), Jember, Indonesia. A total of 600 leaves from 40 Robusta plants were randomly chosen from different growth stages and field’s conditions (managed/irrigated and unmanaged/fully-rainfed plantations). These selected samples were analyzed using different measurement methods (destructive and non-destructive). Variation in weather (natural light intensity) and field’s condition (shade trees) could significantly affect quality of the results of above-canopy measurement. Due to these problems, two methods of assessing plant biophysical properties under varied shade trees were explored, namely direct-leaves and above-canopy measurements. Direct-leaves measurement with custom small chamber was proposed and evaluated to avoid the effect of uncertain weather, shade trees conditions and different time of measurement. The results of this method are encouragingly similar with SPAD-502 Chlorophyll meter and spectrometer, where the coefficient of determination (R2 ) and root mean square error (RMSE) of digital camera, SPAD-502 Chlorophyll meter and spectrometer are 0.7-0.8 and 0.24-0.3 %, 0.68-0.82 and 0.23-0.31%, 0.72-0.85 and 0.21-.29% respectively. While in above-canopy measurement, we propose and evaluate a technique to eliminate variations of light illumination under different shade trees conditions. Incorporating exposure value (EV) with particular vegetation indices (VIs) of each image could improve the R2 in estimating N content of Robusta coffee. We concluded that both the direct-leaves and above canopy measurements using visible cameras performed well in assessing N critical level of Robusta coffee plant. For further image analysis purpose, a web-based integrated coffee management system was designed and developed for analyzing the images obtained from direct-leaf field measurements. Geodatabase was designed and integrated with the prototype of web-based decision support system. This web-based sensing system enabled the farmers / smallholders in uploading the images for remote analysis, and automatically provided the appropriate recommendations for N fertilizer application to their individual plants. For the recommendation, scaling up the system is suggested for accelerating technology transfer to the farmers/smallholders. |
| Year | 2017 |
| Type | Dissertation |
| School | School of Environment, Resources, and Development |
| Department | Department of Food, Agriculture and Natural Resources (Former title: Department of Food Agriculture, and BioResources (DFAB)) |
| Academic Program/FoS | Agricultural Systems and Engineering (ASE) |
| Chairperson(s) | Soni, Peeyush |
| Examination Committee(s) | Sarawut Ninsawat;Loc, Thai Nguyen |
| Scholarship Donor(s) | Directorate General of Higher Education (DIKTI) / Ministry of Research, Technology and Higher Education, Indonesia |
| Degree | Thesis (Ph.D.) - Asian Institute of Technology, 2017 |