Abstract

Changes in water quality in rivers and lakes caused by water pollution are generally investigated by extracting water samples directly from several locations. However, this method is not appropriate for understanding water quality conditions over a large area. On the other hand, remote sensing techniques are especially suitable for understanding water quality conditions over a large area. In our previous studies, we applied fuzzy c-means (FCM) clustering for analysis of water quality parameters in Lake Hachiroko. The results indicated that FCM was effective in estimating the level of pollution in Lake Hachiroko. Moreover, we considered the setting of the best initial values in FCM. The result showed that estimation maps by FCM showed detailed water quality conditions in Lake Hachiroko. However, some estimation maps were classified at the same level in the whole region because the slice level range preset was too wide. The estimation maps created by FCM were in agreement with the water quality conditions, while identifying the water pollution area in the lake was difficult. In this study, we proposed stepwise level slice processing to create water quality estimation maps by FCM. As a result, FCM with stepwise level slice processing (i) creates water quality estimation maps that identify the pollution area and (ii) is useful for water quality estimation of Lake Hachiroko.

Highlights

  • Changes in water quality in rivers and lakes caused by water pollution are generally investigated by extracting water samples directly from several locations

  • To create the estimation map according to the water quality levels in the lake, we studied the setting of the initial values in fuzzy c-means (FCM) using water quality data synchronized with observation times of remote sensing data

  • We proposed stepwise level slice processing to create water quality estimation maps by FCM

Read more

Summary

Introduction

Changes in water quality in rivers and lakes caused by water pollution are generally investigated by extracting water samples directly from several locations This method can obtain only local information and is not appropriate for understanding water quality conditions over a large area. We applied fuzzy c-means (FCM) clustering for analysis of water quality parameters in Lake Hachiroko(2-4) This method considers the effects of specific disturbances and uncertainties on remote sensing data. To create the estimation map according to the water quality levels in the lake, we studied the setting of the initial values in FCM using water quality data synchronized with observation times of remote sensing data. The estimation maps by FCM with increasing slice levels are difficult to decipher because of too much detail depending on the water quality data. To evaluate the usefulness of FCM, we compared the estimation maps with the water quality data

Study Area
ASTER Data
Water Quality Measurements
Process for Analysis
The Initial Values Setting in FCM
Stepwise Level Slice Processing
C1: Cent er of Class 1 C2
Evaluation Criteria of the Estimation Maps by FCM
Estimation Maps by Fuzzy C-Means Clustering
Analysis of Water Quality Conditions
Conclusions
Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call