This research is one of a kind which encompasses the optical and microwave sensor to detect Ganoderma disease in oil palm. The advantage of this data fusion would offer spectral and structural information of the oil palm plantation. A more comprehensive decision can be made for aerial mapping of the oil palm plantation once the model of the data fusion is established and verified.
This research will benefit the scientific community (particularly in the remote sensing field) as the remote sensing of oil palm is a relatively new field. It will also be of benefit to the agricultural community in terms of providing the necessary information required for decision making and management tasks in disease control. This research can be a reference to other sectors that are interested in detecting disease of any plantation and could be tailor-made to suit other vegetation and diseases.
It is widely recognized that image-based remote sensing can provide spatially and temporally distributed information on soil and crop characteristics including tillage and evapotranspiration (ET) from plot to regional scales. ET is an important component of the water balance and the major consumptive use of irrigation water and precipitation on cropland. Numerous ET models have been developed in the last three decades to make use of visible, near‐infrared (NIR), shortwave infrared (SWIR), and most importantly, thermal data acquired by sensors on airborne and satellite platforms. However, there are numerous challenges related to spatial and temporal resolutions of remotely sensed data. These include the need for a comprehensive database for model development, enhancement, and testing; availability of cloud-free images; high quality weather data; accurate estimation and partitioning of evaporation and transpiration; and user-friendly platforms for distributing ET maps for irrigation scheduling and groundwater management purposes in arid and semiarid regions such as the Texas High Plains. In the last decade (2005-2015), researchers at the USDA-ARS Conservation and Production Research Laboratory, home of four large monolithic weighing lysimeters, in collaboration with Texas A&M AgriLife Research and Extension and other research institutions in the U.S., have conducted extensive research on ET remote sensing to address the aforementioned challenges. Research accomplishments include
development
of a comprehensive database consisting of high spatial and temporal resolution remote sensing datasets; measurements of turbulent fluxes with eddy covariance and Bowen ratio systems, sensible heat flux induced light intensity scintillations, net radiation and soil heat fluxes, crop growth parameters, and ET by mass balance. This has allowed the development, enhancement, and testing of numerous remote sensing ET models; development of a framework to generate daily time series ET maps; and development of ET mapping and reference ET software. In this paper, we discuss these accomplishments and the status of ongoing research projects in detail.
The initial study was to evaluate the potential of hyperspectral remote sensing in early detection (prior to visible leaf symptom) in oil palm in three experimental scales: the nursery, field and airborne. The first part of the study employed the nursery datasets using artificial
inoculation to artificially infect healthy oil palm seedlings with the BSR disease and to study the effects and relationships with pathological, physiological and spectral properties for a better understanding of the behaviour of the disease if possible before it appears to the naked eye. The experiments were also designed to determine the stimulate effect of oil palm age in relation to the behaviour of the infection and changes on spectral properties. The second part
of the study
planted palm spectra in the oil palm plantation to evaluate
the spectral responses
of
investigated the spectral responses to Ganoderma
BSR. The ground data in this chapter are directly related to the standing
palm
spectra
in the AISA image pixel used in subsequent analysis. In the final part
of the study, the AISA Eagle hyperspectral datasets with
resolution of
0.68 m and spectral 244 bands were
and
analysis on
the
standing
evaluated.