GEO DATA

Original Papers

Improvement of Algal Bloom Identification Using Satellite Images by the Algal Spatial Monitoring and Machine Learning Analysis in a New Dam Reservoir: Hye-Suk Yi, Sunghwa Choi, Dong-Kyun Kim, Hojoon Kim; GEO DATA. 2023;5(3):126-136. Published online September 25, 2023; DOI: https://doi.org/10.22761/GD.2023.0021

1,023 View
59 Download
1 Citations

Algal blooms are major issues and an ongoing cause of water quality problems in inland waters globally. In the case of harmful algal blooms, the water temperature rises after nitrogen and phosphorus inflow, which occurs in the summer, is the main cause of the algae bloom. In South Korea, algae monitoring methods have been performed by collecting water in point monitoring stations. Recently, in order to overcome the limitations of these existing monitoring methods, spatial monitoring methods using hyperspectral images and satellite images has been researched. We used satellite images for analysis of the spatial algal variation. The accuracy of algal identification is imperative for effective spatial monitoring of algal blooms in the context of ecological health and assessment. In this study, we generated algal big-data with simultaneously observed chlorophyll-a concentrations based on fluorescence measurement and predicted chlorophyll-a concentrations using 13- band satellite images derived from Sentinel-2. In order to validate the values from the satellite images, we compared them with simultaneously observed chlorophyll-a concentrations based on fluorescence measurement. The goal of this study is to improve the accuracy of predictions induced from satellite images. The analytical techniques were comparatively evaluated. The results showed that Artificial Neural Networks exhibited the best performance among them, improving more than 30% accuracy compared to that of multiple linear regression. Furthermore, the accuracy of identifying algal blooms has been shown to increase at high algal concentrations. In the end, it was successful to create algal bloom maps using a new algorithm to analyze algal bloom management.

Citations

Citations to this article as recorded by

Assessment of the Usability of the Linkage between GLORIA and Sentinel-2 Imagery for the Surveillance of Algal Blooms in Freshwater Ecosystems
Gibeom Nam, Sunghwa Choi, Euiho Hwang, Kimook Kang, JinGyeom Kim, DongHyeon Yoon
GEO DATA.2024; 6(4): 451. CrossRef

A Study on C-band Synthetic Aperture Radar Soil Moisture Estimation Based on Machine Learning Using Soil Physics, Topography, and Hydrological Information: Jeehun Chung, Yonggwan Lee, Jinuk Kim, Wonjin Jang, Seongjoon Kim; GEO DATA. 2023;5(3):137-146. Published online September 22, 2023; DOI: https://doi.org/10.22761/GD.2023.0026

1,006 View
52 Download

Abstract PDF: In this study, we applied machine learning to estimate soil moisture levels in South Korea by harnessing data from the Sentinel-1 C-band synthetic aperture radar (SAR). Our approach incorporated not only the relationship between backscattering coefficients and soil moisture but also diverse physical characteristics. This encompassed topographic information, soil physics data, and antecedent precipitation which is a hydrological factor influencing the initial condition of soil moisture. We applied a variety of machine-learning techniques and conducted a comprehensive analysis to compare the performance of each model.

Comparative Study of Machine Learning and Deep Learning Models Applied to Data Preprocessing Methods for Dam Inflow Prediction: Youngsik Jo, Kwansue Jung; GEO DATA. 2023;5(2):92-102. Published online June 30, 2023; DOI: https://doi.org/10.22761/GD.2023.0016

1,206 View
56 Download

Abstract PDF: In this study, we employed representative machine learning (ML) and deep learning (DL) models previously utilized in the fields of rainfall and runoff analysis in the water resources sector. We not only performed hyperparameter tuning of the models but also considered the characteristics of the model and the combination and preprocessing (such as lag-time and moving average) of meteorological and hydrological data. We then compared and evaluated the performance of the models according to various scenarios of data characteristics and ML & DL model combinations for predicting daily water inflow. To accomplish this, we utilized meteorological and hydrological data collected from 1974 to 2021 in the Soyang River Dam Basin to examine 1) precipitation, 2) inflow, and 3) meteorological data as primary independent variables. We then employed a total of 36 scenario combinations as input data for ML & DL, applying a) lag-time, b) moving average, and c) component separation conditions for inflow. To identify the most suitable data combination characteristics and ML & DL models for predicting daily inflow, we compared and evaluated 10 different ML & DL models: 1) Linear Regression, 2) Lasso, 3) Ridge, 4) Support Vector Regression, 5) Random Forest (RF), 6) Light Gradient Boosting Model, 7) XGBoost for ML, and 8) Long Short-Term Memory (LSTM) models, 9) Temporal Convolutional Network (TCN), and 10) LSTM-TCN for DL.

Article

Reconstruction of monthly fCO₂ distribution in the Ross Sea, Antarctica during 1998 -2018 using machine learning technique and observational data sets: Ahra Mo, Jung-ok Choi, Keyhong Park; GEO DATA. 2022;4(3):15-24. Published online September 30, 2022; DOI: https://doi.org/10.22761/DJ2022.4.3.003

916 View
27 Download

Abstract PDF: The ocean is a major reservoir of anthropogenic carbon dioxide, especially the Southern Ocean has been known to absorb 40% of the carbon dioxide emitted by human activity. The Ross Sea is one of the most productive regions in the Southern Ocean; however, its carbon dioxide absorption capacity has not been clearly evaluated yet. Because the Southern Ocean is geographically isolated from civilization and thus, its remoteness prevents making sufficient observations from proving reliable carbon dioxide sink strength estimates. Thus, in order to overcome the current spatial and temporal limitations of direct observations, the fugacity of carbon dioxide (fCO₂) data was reproduced using a machine learning technique (i.e., random forest technique). The technique is a type of machine learning frequently used to reproduce marine environmental variations through training satellite data and modeled data as well as existing observational data. Furthermore, to reproduce more reliable fCO₂ estimates, in addition to marine environmental variables (i.e., sea surface temperature, sea ice concentration, and chlorophyll-a concentration), cloud cover, wind speed, and El Niño index were included in the machine learning procedure. In this study, we provide the past 21 years (1998 – 2018) of monthly spatial and temporal variation information of dissolved carbon dioxide in the Ross Sea, Antarctica.

First
Prev
Page of 1
Next
Last

Search