RIASSUNTO
With the rising global ocean temperatures, the habitats of Marine organisms change, which affecting human production practices at the same time. First of all, in order to explore the relationship between global ocean temperature and time, we collected sea surface temperature (SST) data in the north Atlantic region. We establish the BP neural network prediction model based on EMD. The prediction model firstly used EMD to stabilize the water temperature time series, and obtained a group of stable components IMF and a surplus. Then the BP neural network is used to predict each component, and the predicted value is added as the predicted value of the original sequence. To get a forecast for the sea surface temperature of Scotland over the next 50 years. Visualize the past and predicted the data and match the range of suitable temperatures for herring and mackerel to the sea surface temperatures near Scotland. Matching area represents the fish habitat. Observing the movement of the matching area each year, the fish's habitat migration process can be predicted. The second problem is to choose the suitable operating strategy for fishing company. Because the migration of fish is sensitive to the changes in temperature, there will be a risk of secondary transfer if choosing transfer assert. Thus, we suggest to use small fishing vessels.