Modeling Storm Tides: Integrating Physical Parameters and Machine Learning

By Jurassic JennJun 24, 2024 14:39 PMTech
Share:
Modeling Storm Tides: Integrating Physical Parameters and Machine Learning
Modeling Storm Tides: Integrating Physical Parameters and Machine Learning. Source: techxplore.com

Predicting extreme weather events is crucial in safeguarding vulnerable regions, especially in the face of climate change. Santos, the largest port in Latin America located on the coast of São Paulo state in Brazil, has been the subject of extensive studies due to the threat posed by storm surges to its infrastructure and local ecosystems. This was reported by SSPDaily.

Recently, a noteworthy study focusing on a critical area in Santos utilized advanced machine learning techniques to enhance existing systems for predicting extreme weather events. Published in the Proceedings of the AAAI Conference on Artificial Intelligence, this research effort involved a collaboration among numerous researchers and was coordinated by Anna Helena Reali Costa, a distinguished professor at the University of São Paulo's Engineering School (POLI-USP). The primary author of the study is Marcel Barros, a researcher in the Department of Computer Engineering and Digital Systems at POLI-USP.

Traditionally, models used to predict variables such as sea surface heights, high tides, and wave heights are based on complex differential equations that incorporate temporal and spatial information, including astronomic tide, wind regime, current velocity, and salinity. These models have achieved success in various areas; however, they rely on simplifications and assumptions, not easily accommodating new data sources to enhance forecast reliability.

To overcome these limitations, the research team fused the fields of physical modeling and machine learning. Their model, classified as physics-informed machine learning (PIML), leverages physical models as a foundation but leverages measured data to refine predictions. In essence, the PIML approach harmonizes these two sources of information to develop forecasts that are both precise and accurate.

A significant challenge arose when incorporating sensor data into the model due to its irregular nature and issues such as missing data, temporal displacements, and variations in sampling frequencies. The research team tackled this obstacle by innovatively representing the passage of time in neural networks. This novel representation allows the model to account for missing data windows, ultimately yielding more accurate predictions for tide and wave heights.

This innovative technique not only accomplishes improved modeling of complex natural phenomena, such as storm tides but also holds promise for applying irregular time series analysis to other fields, including healthcare, sensor networks in manufacturing, and financial indicators.

Moreover, the model embraces a multimodal architecture that combines different types of neural networks. Satellite images, tables, and forecasts from numerical models are already integrated successfully, with future potential for incorporating additional data types like text and audio. This multimodal approach signifies a significant stride towards robust and adaptable forecasting systems adept at handling the intricacies and fluctuations of data associated with extreme weather events.

There are three key virtues to this newly developed model, as articulated by Reali Costa. Firstly, it bridges the gap between physical and numerical models. Secondly, it presents a fresh perspective on representing time in neural networks. Lastly, it demonstrates versatility in processing data of various formats through its multimodal architecture.

Reali Costa emphasizes the implications of this study beyond the realm of storm tide predictions in Santos. The proposed methodology has the potential to enhance the accuracy of forecasts for other extreme weather events. Simultaneously, it illuminates the challenges involved in integrating physical models and sensor data in complex contexts, laying the groundwork for potential solutions.

In conclusion, the fusion of physical parameters and machine learning in the modeling of storm tides sets a new benchmark for predicting extreme weather events. This research showcases the power of blending advanced technologies to optimize forecasting systems and exemplifies the opportunities that lie ahead in developing more reliable and effective strategies to safeguard vulnerable regions worldwide.

Top Articles
To see a white owl - meaning

The spiritual significance of seeing a white owl in dreams and in reality: symbolism of the totem

Jul 17, 2024 18:20 PM
What does the name Austin mean

The meaning of the name Austin in spiritual terms: what the letters symbolize

Jul 17, 2024 14:13 PM

The meaning behind the Angel number 3: spiritual insights

Jul 17, 2024 11:32 AM

Symbolism of seeing snakes & spiritual meaning of meeting

Jul 12, 2024 16:59 PM
More News
A surge of confidence and charisma in proffesional life: horoscope for July 20-22

A surge of confidence and charisma in proffesional life: horoscope for July 20-22

Jul 20, 2024 12:41 PM
 Three zodiac signs are aligned for unexpected financial gains soon

Three zodiac signs are aligned for unexpected financial gains soon

Jul 20, 2024 12:21 PM

Michael Jordan had a surprising superstition he adhered to before every game

Jul 20, 2024 11:39 AM

AC Milan Bolster Attack with Alvaro Morata Signing from Atletico Madrid

Jul 20, 2024 11:17 AM
Three zodiac signs to welcome ubiquitous joy: horoscope for July 20

Three zodiac signs to welcome ubiquitous joy: horoscope for July 20

Jul 20, 2024 10:58 AM
The most anticipated TV shows of summer 2024

The most anticipated TV shows of summer 2024

Jul 20, 2024 10:35 AM
Elvis Presley

10 Fascinating Insights About Elvis Presley

Jul 20, 2024 10:13 AM
Intense solar flares predicted to hit Earth this weekend

Intense solar flares predicted to hit Earth this weekend

Jul 20, 2024 09:52 AM
Laughter at Woman's Reaction When Wife Asks to Split Bill

Laughter Erupts as Wife Stuns Husband with Suggestion to Split Bill: A Hilarious Moment Captured

Jul 20, 2024 09:32 AM

Curried sausages: a budget-friendly delicacy for the whole family

Jul 20, 2024 09:09 AM

This simple hack will help you clean greasy food containers without much effort

Jul 20, 2024 08:48 AM
Three zodiac signs most likely to land their dream job in the upcoming months

Three zodiac signs most likely to land their dream job in the upcoming months

Jul 20, 2024 08:26 AM
Google's shortened links to cease functioning by August 2025

Google's shortened links to cease functioning by August 2025

Jul 20, 2024 08:05 AM
Banks, airlines, and other major businesses worldwide flooded with Window's 'Blue Screen of Death' after biggest IT outage ever

Banks, airlines, and other major businesses worldwide flooded with Window's 'Blue Screen of Death' after biggest IT outage ever

Jul 20, 2024 07:44 AM
Three zodiac signs will be pleased with positive interpersonal connections: horoscope for July 20

Three zodiac signs will be pleased with positive interpersonal connections: horoscope for July 20

Jul 20, 2024 07:23 AM

Seeking out new educational opportunities: horoscope for July 20-21

Jul 20, 2024 07:04 AM

Three zodiac signs most likely to encounter tranquility during turbulent times

Jul 19, 2024 21:59 PM
Three zodiac signs will get a chance to make steady progress in their work

Three zodiac signs will get a chance to make steady progress in their work

Jul 19, 2024 21:35 PM
"I got ultimate revenge on my lazy colleague – and my boss backed me"

"I got ultimate revenge on my lazy colleague – and my boss backed me"

Jul 19, 2024 21:17 PM
Humans reached Argentina by 20,000 years ago

Humans reached Argentina by 20,000 years ago — and they may have survived by eating giant armadillos, study suggests

Jul 19, 2024 20:56 PM
10 Astonishing Facts About Monica Bellucci

10 Astonishing Facts About Monica Bellucci

Jul 19, 2024 20:34 PM
Cavan Sullivan

Philadelphia Union Midfielder, Cavan Sullivan, Breaks MLS Record as Youngest MLS Player

Jul 19, 2024 20:12 PM

Curry chicken salad with grapes: a delightful ensemble

Jul 19, 2024 19:52 PM

Microsoft's Hiring of Inflection AI Staff Under Investigation by UK Regulatory Authority

Jul 19, 2024 19:29 PM
The three zodiac signs are most likely to find unexpected tranquility in their personal lives

The three zodiac signs are most likely to find unexpected tranquility in their personal lives

Jul 19, 2024 19:08 PM
The Top 5 Gary Oldman Movies Ranked

The Top 5 Gary Oldman Movies Ranked

Jul 19, 2024 18:46 PM