Life Preserving Power with DNR's Artificial Neural Network
The Division of Nearshore Research has had great success in forecasting water level along the Texas coast. Several models have been developed to accomplish this task. Water level measurements and water level forecasts created by dynamic models such as harmonic analysis (tide tables) and persistence models are presented in the graph below for the Bird Island location.
| + Observed water level | _ Tide chart predictions | _ Persistence model predictions |
In this graph, the black "+" signs represent the observed water level values, blue represents water level predictions through tide charts, and green represents water level predictions through the persistence model. An Artificial Neural Network (ANN) model will soon be implemented for this station as well.
An ANN may also be trained to forecast other important values. DNR is in the process of training an ANN to predict water temperatures within the Laguna Madre. The purpose of the ANN water temperature forecasting is to alert us to when temperatures will most likely force aquatic life to congregate in the deeper waters of the Gulf Intra-Coastal Water Way. Being able to forecast harsh temperatures in the water column will be vital to the preservation of fish species of the Laguna Madre. Soon, we will be able to deliver water temperature forecasts just as we forecast water level, as shown in the graph above.
Contact Information: amostella@lighthouse.tamucc.edu
Texas A&M University-Corpus Christi