IOTC Area Predictive analysis

Description

Hypothesis and Thesis

Approach

Our first attempt aimed to discover hidden periodicities in the exploitation of the ocean area. Thus, we traced the trend of the fishing hours effort in the period between January 1997 and December 2012. We summed the hours spent in the same month and eventually obtained the trend in figure 1.

Fishing hours aggregated per month along the entire period January 1997-December 2012

Furthermore, we associated the fishing hours in one single month to the barycenter of the areas interested by the fishing activity in that month. The monthly barycenters were calculated by weighting each 1 degree cell in the original dataset according to the fishing hours spent in that month in the cell. The resulting dataset can be found here. This contains unfolded information about the barycenters coordinates and their distance from the origin of the reference system. Figure 2 and 3 display the trends for the longitudes and latitudes of the barycenter. The signals show a periodic behaviour, which we investigated more accurately. In figures 4 and 5 we report the trends for the amount of catch in tonnes of the Yellowfin Tuna and the related catch per unit of effort (CPUE) calculated as the ratio between the catch and the fishing hours.

Distribution of the vessels longitudes for Yellofin Tuna by Purse Seine technique, in the entire period January 1997-December 2012

Distribution of the vessels latitudes for Yellofin Tuna by Purse Seine technique, in the entire period January 1997-December 2012

Catch quantity in tonnes for Yellofin Tuna, in the entire period January 1997-December 2012

CPUE for Yellofin Tuna, in the entire period January 1997-December 2012

Spectrogram for the longitude trend.

Spectrogram for the latitude trend.

Longitude trend in samples.

Latitude trend in samples.

In order to inspect the periodic behaviour of the trends of the coordinates, we applied Fourier analysis to the signals and traced the spectrograms. In both the cases, the analysis detected a period of in the samples, which corresponds to a periodic exploitation of some areas.

Fishing hours trend in samples.

Spectrogram for the fishing hours trend.

Fishing effort, instead, is not a stationary signal, thus we tried to inspect its hidden periodicities by means of a Short-Time Fourier analysis. The analysis was made on chunks of the signal with quite stationary mean and variance in the samples values.

Distribution of the presence of vessels on 1 degree csquares, in the entire period January 1997-December 2012

We also report the trend of the Csquare codes associated to the barycenters. Each code was assigned to an index and the chart displays the distribution of such indexes in time. This is one way to report a bidimensional occurrence on one dimension.The trend of the indexes is periodic and this is due to the periodic trend in longitudes and latitudes.

Forecasted trend for latitude.

Forecasted trend for longitude.

Finally, we tried to forecast the coordinates trends in order to understand which places could be interested by exploitation in the 12 next months. To such aim we used Singular Spectrum Analysis.

Fishing hours trend in samples.

Spectrogram for the fishing hours trend.

Forecasted trend for fishing hours.

We applied the same analysis to the fishing hours trend. The analysis detected a weak periodicity of 12.91 samples with indecision interval between 13.25 and 12.59. Furthermore, the forecast suggests that the amount of hours per month is going to oscillating aroung lower values in average, with respect to the previous years.