Thread by @aj_wahyudi

Multi-temporal variability forecast of particulate organic carbon in the Indonesian seas

#forecasting
#Python
#MachineLearning
#Oceanography
#IndonesianSeas
#POC
#biogeochemistry

@IMBeRgroup_RCO

link.springer.com/article/10.100…

This research aimed to forecast the temporal variability of POC in Indonesian waters.

POC forecasting is essential in predicting the productivity of Indonesian seas, for several reasons, namely, that it will affect marine productivity, the Indonesian seas are dynamic, and the existence of global threats to the ocean (climate change, ocean acidification, and d'Ox)

We used the Seasonal Autoregressive Integrated Moving Average (SARIMA) statistical learning forecasting model. There is no previous study has examined forecasts of the POC level using this approach.

The hindcast estimates of POC in the Indonesian seas show a decreasing tendency. A slight decreasing tendency is also shown by the forecasting monthly data
series using the cycle frequency s=60.

The decreasing tendency of POC may be due to the effect of warming and ocean acidification on phytoplankton.

Furthermore, a decrease in the decomposition rate will potentially lead to a decrease in POC net concentration.

Inter-annual forcing factors (e.g., IOD and ENSO) also affect the dynamic of POC. High POC peaks are forecasted in 2024, 2027,
and 2029.

Since POC temporal variability appears to be affected by physical, chemical, and climatic drivers, it may be insufficient to rely on the SARIMA for forecasting.

The present study thus emphasizes that the viable type of forecasting may involve the use of monthly data series with a multi-annual scenario (s=60).

Therefore, adding exogenous factors such as the Dipole Mode Index (for IOD) or Nino3 index (for ENSO) could increase the
robustness of the forecasting.

In addition, the POC dynamic
is affected by sea surface temperature, terrestrial input, primary production, chlorophyll-a, and salinity change. These could also be treated as exogenous factors.

A future study could thus add an exogenous factor to SARIMA so that it becomes a Seasonal Autoregressive Integrated Moving Average with eXogenous factors/SARIMAX model.

This tweets are attributed to Wahyudi et al., 2023 (and the reference there in).

@threadreaderapp unroll