Scholarlink

This study applies Function-on-Function Linear Quantile Regression (FFQR) to assess the influence of key marine variables, salinity, conductivity, chlorophyll, pH, and turbidity, on marine water temperature along the Abruzzo coastline (Italy). Unlike traditional regression models focused on mean trends, FFQR allows for a detailed analysis across different quantiles, capturing variability and extreme values associated with thermal stress. Functional data collected at multiple depths and transects are modeled using a combination of Functional Principal Component Regression and quantile techniques. The results show that conductivity, chlorophyll (a), pH, and turbidity significantly affect the upper quantiles of the temperature distribution, supporting the utility of FFQR in detecting marine heat anomalies and revealing complex ecological patterns.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Quantile Functional Regression Approach to Marine Data Analysis

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Annalina Sarra,
  • Adelia Evangelista,
  • Tonio Di Battista,
  • Nicola Di Deo

摘要

This study applies Function-on-Function Linear Quantile Regression (FFQR) to assess the influence of key marine variables, salinity, conductivity, chlorophyll, pH, and turbidity, on marine water temperature along the Abruzzo coastline (Italy). Unlike traditional regression models focused on mean trends, FFQR allows for a detailed analysis across different quantiles, capturing variability and extreme values associated with thermal stress. Functional data collected at multiple depths and transects are modeled using a combination of Functional Principal Component Regression and quantile techniques. The results show that conductivity, chlorophyll (a), pH, and turbidity significantly affect the upper quantiles of the temperature distribution, supporting the utility of FFQR in detecting marine heat anomalies and revealing complex ecological patterns.