spot_img
26.3 C
Philippines
Friday, January 24, 2025
26.3 C
Philippines
Friday, January 24, 2025

UP Diliman experts determine optimal model for predicting lake algal blooms

Estimated reading time: 1 minute and 26 seconds
16px

Algal blooms, which can turn water green and kill fish, pose a significant threat to Laguna Lake, a major source of food and drinking water for Metro Manila.

Researchers from the University of the Philippines Diliman (UPD) have developed a new system to predict algal blooms in the lake using machine learning (ML) models.

- Advertisement -

Algal blooms occur when algae overgrow in bodies of water. A standard method for monitoring algal populations is to measure chlorophyll-a, a green pigment produced by algae. However, this method often detects blooms after they have already occurred.

The UPD researchers, led by Dr. Karl Ezra Pilario of the Department of Chemical Engineering, found that monitoring nitrate and phosphate concentrations in the water, which are often linked to increases in chlorophyll-a, is a more effective approach.

They compared the performance of eight common ML models using water quality data from Laguna Lake and historical data from global lakes.

Two models, Kernel Ridge Regression (KRR) and Gaussian Process Regression (GPR), were found to be the most accurate and robust, effectively handling noisy data.

“Now that we have an accurate, robust, and explainable predictor of chlorophyll-a, we can deploy the model for rapid detection of impending algal blooms,” Pilario said in a statement.

The researchers recommend monthly monitoring of nitrate and phosphate levels in the lake to allow for timely interventions to mitigate the impact of potential blooms.

They are now exploring ways to improve the models by incorporating additional predictors such as weather conditions and land cover types.

“In the future, we encourage researchers to test for the robustness and explainability of their machine learning models, and not just for accuracy, to enhance the credibility of the results for policy-making,” Pilario said.

LATEST NEWS

Popular Articles