Keywords

grain-size distributions, Fréchet distance, environmental monitoring.

Location

Session C1: VI Data Mining for Environmental Sciences Session

Start Date

13-7-2016 9:50 AM

End Date

13-7-2016 10:10 AM

Abstract

Grain-size is the most fundamental property of sediment particles, its analysis is essential to understand sediment provenance, transport history and depositional conditions. In the ambit of environmental monitoring a key point is often the study of the variations of sediments composition in terms of variation of their grain-size distributions. In this work it is presented a procedure based on the computing of easy mathematical indexes useful to perform in an effective way the comparison of grain-size distributions. The developed procedure consists in two steps of analysis to quantify the dissimilarities between grain-size distributions and to characterize the typology of occurred variations.. A validation process is executed to verify the proper work of the procedure, using a large dataset of grain-size distributions.

Two possible applications of the procedure are presented, one to study spatial alterations of sediments composition and another one to analyze temporal changes of sediment in the same sampling station.

The proposed procedure allows to analyze in a quick way large datasets and it is flexible tool to be adapted to the peculiarities of the analyzed data, in order to optimize the achievable results.

COinS
 
Jul 13th, 9:50 AM Jul 13th, 10:10 AM

A Mathematical Procedure To Estimate Variations Between Grain-Size Distributions

Session C1: VI Data Mining for Environmental Sciences Session

Grain-size is the most fundamental property of sediment particles, its analysis is essential to understand sediment provenance, transport history and depositional conditions. In the ambit of environmental monitoring a key point is often the study of the variations of sediments composition in terms of variation of their grain-size distributions. In this work it is presented a procedure based on the computing of easy mathematical indexes useful to perform in an effective way the comparison of grain-size distributions. The developed procedure consists in two steps of analysis to quantify the dissimilarities between grain-size distributions and to characterize the typology of occurred variations.. A validation process is executed to verify the proper work of the procedure, using a large dataset of grain-size distributions.

Two possible applications of the procedure are presented, one to study spatial alterations of sediments composition and another one to analyze temporal changes of sediment in the same sampling station.

The proposed procedure allows to analyze in a quick way large datasets and it is flexible tool to be adapted to the peculiarities of the analyzed data, in order to optimize the achievable results.