Hysteresis loop model for the estimation of the coastal water temperatures

Abstract

Water temperature (hereafter WT) is one of the most important a-biotic factors influencing the distribution of marine species. It controls the rate of metabolic and reproductive activities, and determines which species can survive. However, the WT data is very poor in comparison with the air temperature (hereafter AT) data. This situation has been solved by the WT estimation method based on the regression model by using the AT data. This method, however, cannot consider the hysteresis which is the general pattern to be shown between AT and WT data relationship. It has not the time-history concept. In this study, the Hysteresis Loop model based on the harmonic analysis is developed. The new model is novel in that it requires only single variable (air temperature) as an input to convert it to projected water temperature, and that it captures the hysteresis pattern of the rising limb and falling limb and also historic memory. It is able to consider the hysteresis pattern and includes the time-history concept. The model calibration and validation is carried out to check the RMS error by using the 4-year period (2005. 7- 2009. 6) AT and WT data of the monitoring buoy in Mikawa bay. The results show that the RMS error of the WT estimation only using the AT data is in the range of 0.8 ?1.0 at the calibration and validation stage. It is better in comparison with the RMS error 1.4-2.1 of the linear regression model.The Hysteresis Loop model can be used as an efficient tool for the prediction of the present and near future WT in the coastal area. It can be regarded to predict the more accurate and reliable WT information in the condition of the WT rise in the (near) future due to the global warming effects.