VMS-based Trajectory Analyses for the Identification of Longline Vessel Activities

Get Complete Project Material File(s) Now! »

Illegal, Unreported and Unregulated (IUU) Fishing

FAO defines Illegal fishing as fishing activities conducted by national or foreign vessels in waters under the jurisdiction of a State, without the permission of that State, or in contravention to laws and regulations in some other manner. Unreported fishing is defined as fishing activities, which have not been declared, or have been misdeclared, to the relevant national authority, in contravention to national laws and regulations. Unregulated fishing are defined as fishing activities in the area of application of a relevant regional fisheries management organization that are conducted by vessels without nationality, or in areas or for fish stocks in relation to which there are no applicable conservation or management measures [9].
IUU fishing may relate to different causes. It is highly attractive for no taxes on the catches, it is interested in trading limited species with high price, it occurs widely in the high seas [11]. It often employs harmful fishing gear that produces detrimental effects on the environment [5]. IUU fishing threatens overexploitation of fish stocks and could be an obstacle to the recovery of fish population and ecosystems [10]. Agnew et al. estimated the total value losses between $10 bn and $23.5 bn per year globally, in tonnes between 11 and 26 million. They noted that developing countries are more vulnerable to illegal fishing.

IUU Fishing in Indonesia

As a big archipelago country, the maritime borders are widely spread by little islands which are located in hinterland and most outside. Indonesian sea has boundary with ten neighboring countries. The illegal fishing is highly susceptible to happen in Indonesia due to different factors i.e. a large number of fish stocks, the lack of monitoring, control, surveillance at the boundary area, weak governance. Since October 2014, Indonesia has taken harsh action by blowing up the vessel which is doing the illegal fishing in Indonesian territorial waters.
During the period October 2014 – December 2015, the total amount of 121 vessels had been only in EEZ but also in archipelagic waters. The non-compliance to Indonesian laws include: the absence of fishing licenses, the manipulation of vessel’s information, the use of different fishing’s gear from license prohibited, and turning off the VMS transmitter [14]. Illegal fishing by foreign vessels mainly occur in Indonesian EEZ of the South China Sea, the Sulawesi Sea and Arafura sea [14]. In general, unreported fishing in Indonesia relates to the catch production. The others caused include sea transshipment, to directly convey the fish catch to abroad [13]. Many licensed foreign vessels do not report to the fishing port or land the fish catch by sea transshipment, especially in Indonesian EEZ of the South China Sea and in Sulawesi Sea [11].
Unreported fishing in Indonesia often refers to sport fishing [13]. It involves no speciesspecific regulation, particularly ornamental fish, and modified fishing gear [11].
In Indonesia, IUU fishing is having impact on social, economic and environmental aspects and also conflict with traditional fishers [13].
– social aspects include threats on the traditional fishers having the small size of vessels and traditional fishing gears compare to the illegal fishing vessel, reduction of catch production caused by the damage of environment [14].
– economic aspects involve the loss of catch fishes production as well as nationwide loss due to the non-collected tax revenue[13].

READ  Optimal Algorithms for Linear Best-ArmIdentification

Spatial-temporal Matching of VMS Data and Observer Data

In this work we assess the quality of observer data that could be used for validation and training purposes in the proposed framework. We proceed as follows:
– Matching vessel names between the observer data and the VMS database. The result of this step is shown in Fig. 5. Only 20 longliners and 27 purse-seiners from observer data could be retrieved based on their exact name in the VMS dataset. Hence 22 vessels could not be matched to a vessel in the VMS dataset. This might refers to vessels ≤ 30 GT with no requirement for the use of a VMS transmitter.
– Matching space/time positions recorded by the observers to VMS data. For a given vessel, we considered that when 70% or more of the observer data showed a good match to VMS data in terms space-time positions, the observer data was a relevant for validation and training purposes. As reported in Table 8, only 13 % of observer data collected by SDI and LPPT were tagged as good quality data based on this criterion, that is to say only 6 longliners and 4 purse-seiners. We illustrate in Fig. 9 an example of position mismatch between the observer and the VMS dataset.

Table of contents :

hapter Description Page
Acknowledgments
Abstract
Résumé
List of figures
List of tables
Acronyms
Contents
Résumé de la thèse
Ch. 2 Introduction 
Indonesian fisheries, IUU fishing and VMS data
2.1 Indonesian fisheries
2.2 Illegal, Unreported and Unregulated (IUU) Fishing
2.3 IUU fishing in Indonesia
2.4 Vessel Monitoring System
2.5 VMS data
2.6 Observer data
2.7 Spatially-Temporally Matching of VMS Data and Observer Data
2.8 New Avenue to Use VMS Data: State of the Art
References
Ch. 3 VMS-based Trajectory Analyses for the Identification of Longline Vessel Activities
3.1 Introduction
3.2 Material and Methods
3.2.1 Dataset
3.2.2 Instantaneous vs. Calculated Speed and Turning Angle
3.2.3 Proposed Approach and Models
A. Simple rule-based Model
B. Hidden Markov Model (HMM)
C. Support Vector Machine (SVM)
D. Random Forest
3.2.4 Performance evaluation
3.3 Results
3.3.1 Model Performance
3.4 Discussion and Conclusion
References
Ch. 4 Mapping the Fishing Effort of Indonesian Tuna Longliners from VMS data
4.1 Introduction
4.2 Materials and methods
4.2.1 VMS data and observer data
4.2.2 Seapodym model
4.2.3 Modeling and inference of longliners’ activities from VMS data
4.2.4 Estimating longliners’ fishing effort from VMS data
4.3 Results
4.3.1 Mapping fishing effort in bimonthly (2012-2014)
4.3.2 Mapping of fishing effort in seasonally (2012-2014)
4.3.3 The main hotspot area explored yearly (2012-2014)
4.3.4 Comparisons of longliners’ fishing effort distributions with SEAPODYM biomass predictions
4.4 Discussion and Conclusion
References
Fishing Gear Identification from VMS-based Fishing Vessel
Trajectories
5.1 Introduction
5.2 VMS data
5.3 Proposed approach
A Unsupervised characterization of gear-specific fishing
vessel activity from VMS Trajectories
Gaussian-von Mises Mixture Model (GvMMM)
Gaussian Mixture Model VpVt (GMM-VpVt)
ML estimation of mixture model parameters
B VMS-based feature extraction for fishing gear classification
C Supervised Gear Recognition from VMS Trajectories
5.4 Results
A Gear recognition performance
B Detection of abnormal VMS patterns
5.5 Discussion and Conclusion
References 

GET THE COMPLETE PROJECT

Related Posts