Abstract
Processing and forecasting of the informationon fishing ground by satellite imagery: Comparison between satellite images andin-situ fishery data around Bali
Prof. Dr. Susumu Kanno, Prof. Dr. Yasuhiro Sugimori,
Center for Remote Sensing and Ocean Sciences, the University of Udayana, Indonesia
Yasuo Furushima, Japan Agency for Marine-Earth Science and Technology
Email: whitecat@st.rim.or.jp, sugimoriy@nifty.com , furus@jamstec.go.jp
ABSTRACT
In order to improve the method for prediction of tuna fishing ground, the modification of the analysis about satellite altimeter data was made a trial and discussed. Although several attempts has been carrying out to predict effective fishing ground by satellite imagery: sea surface temperature, ocean color, sea surface height, etc., but most of the predictions rely on the decision of operatorfs psycho-physical judgment about the location of fishing ground. Therefore there are no numerical reason for the prediction and the result of predictions sometimes different among each operator. In this study, we focused on the satellite altimeter, TOPEX/POSEIDON to improve the criterion whether the predicted location is good fishing ground or not. TOPEX/POSEIDON altimeter data has been using for fishery targeting horse mackerel, mackerel pike, or tuna sps. In these cases good fishing ground is predicted by the absolute sea surface height or anomary, this is also said by the another word gpsycho-physicalh judgment, because just relative high, low area or their boundary are defined as good fishing ground. We made a trial to utilize the gradient of sea surface height as a signal for fishing ground prediction.
The TOPEX/POSEIDON altimetry data were collected from the archive system on the web developed by Colorado Center for Astrodynamics Research (CCAR), USA. Sea surface height anomary was used in the analysis because of the reason for normalization of the absolute sea surface height in the model area. Fishery data were supplied by local fishing information around Indonesia and hearing information. Fishery information had been converted to ghook rateh which mean how many tuna are hooked in the one hundred hooks equipped in a unit ghaenawah trawling line. Because if the total catch of the tuna is used for the analysis, the error derived from the capacity in fishing for each fishing boat interferes, therefore the use of hook rate is reasonable for correct and normalized data treatment, and it can be referred as CPUE(Catch Per Unit Effort).
Model area was divided into the mesh grids with the interval of one degree about latitude and longitude according to the spatial resolution of fishery data. The gradient of sea surface height is defined and calculated between the neighbor direction which has the maximum gradient. Using above both gradient of sea surface height and hook rate, the relation ships between them were analyzed over the various location corresponding to in-situ fishing ground. Results showed that the fishery data with hook rate over 0.8 are grouped in a zone from 1.0E-06 to 2.0E-06 of sea surface height gradient and the standard deviation of the data is lower than the case of hook rate up to 0.8. These indicate that this method has not only possibility for the prediction of fishing ground quantitatively but also reasonable accuracy as shown in the change in the standard deviation.
This method can be utilized for the effective fishing plan with the resource protection and the economy in the fishing operation, because if fisherman can predict the fishing ground reasonably the overfishing in particular local area will be avoided.