CoClime - Probability of phytoplankton (“Pseudo-nitzschia seriata” complex ) presence, Southwest Ireland - future prediction (2017 - 2035)

This is the output of GBM (Gradient Boosting Model) analyses of P. seriata presence and absnece data done by MI OCIS and CoCliME project. This output dataset shows prediction of probability of P. seriata presence in the future (2017 - 2035) in SW Ireland. The analyses was performed in R 3.6.3, with the packages tidyverse 1.3.0 for data handling and visualisation, and xgboost 1.2.0.1 for boosted regression analyses. The dataset used for analyses is available on MI data catalogue (http://data.marine.ie/geonetwork/srv/eng/catalog.search#/metadata/ie.marine.data:dataset.4445) and the dataset used for prediction is requested to be included in MI data catalogue ([20]. CDF-t Climate run (Climate_run_corr.rds) in CoCliME model SOP). This dataset is used to visualise the prediction on R Shiny application for present period (https://marine-institute-ireland.shinyapps.io/P_seriata_probability/). This dataset is used for displaying future prediction of P. seriata probability of presence on the Shiny app.

本数据集为MI OCIS与CoCliME项目针对P. seriata存在与缺失数据开展GBM（梯度提升树模型，Gradient Boosting Model）分析所得的输出结果，展示了爱尔兰西南部地区2017年至2035年期间P. seriata存在概率的预测结果。本次分析基于R 3.6.3版本开展，数据处理与可视化环节使用了tidyverse 1.3.0包，梯度提升回归分析则借助xgboost 1.2.0.1包完成。用于本次分析的数据集可在MI数据目录中获取（访问链接：http://data.marine.ie/geonetwork/srv/eng/catalog.search#/metadata/ie.marine.data:dataset.4445）；用于预测的数据集已申请收录至MI数据目录（参考文献[20]：CoCliME模型标准操作程序（SOP）中的CDF-t气候运行数据集Climate_run_corr.rds）。本数据集可用于在R Shiny应用中可视化当前时段的P. seriata存在概率预测结果（访问链接：https://marine-institute-ireland.shinyapps.io/P_seriata_probability/），同时用于在该应用中展示P. seriata未来存在概率的预测结果。