ANN-Based Estimation of Low-Latitude Monthly Ocean Latent Heat Flux by Ensemble Satellite and Reanalysis Products

Abstract

Ocean latent heat flux (LHF) is an essential variable for air–sea interactions, which establishes the link between energy balance, water and carbon cycle. The low-latitude ocean is the main heat source of the global ocean and has a great influence on global climate change and energy transmission. Thus, an accuracy estimation of high-resolution ocean LHF over low-latitude area is vital to the understanding of energy and water cycle, and it remains a challenge. To reduce the uncertainties of individual LHF products over low-latitude areas, four machine learning (ML) methods (Artificial Neutral Network (ANN), Random forest (RF), Bayesian Ridge regression and Random Sample Consensus (RANSAC) regression) were applied to estimate low-latitude monthly ocean LHF by using two satellite products (JOFURO-3 and GSSTF-3) and two reanalysis products (MERRA-2 and ERA-I). We validated the estimated ocean LHF using 115 widely distributed buoy sites from three buoy site arrays (TAO, PIRATA and RAMA). The validation results demonstrate that the performance of LHF estimations derived from the ML methods (including ANN, RF, BR and RANSAC) were significantly better than individual LHF products, indicated by R2 increasing by 3.7–46.4%. Among them, the LHF estimation using the ANN method increased the R2 of the four-individual ocean LHF products (ranging from 0.56 to 0.79) to 0.88 and decreased the RMSE (ranging from 19.1 to 37.5) to 11 W m-2. Compared to three other ML methods (RF, BR and RANSAC), ANN method exhibited the best performance according to the validation results. The results of relative uncertainty analysis using the triangle cornered hat (TCH) method show that the ensemble LHF product using ML methods has lower relative uncertainty than individual LHF product in most area. The ANN was employed to implement the mapping of annual average ocean LHF over low-latitude at a spatial resolution of 0.25° during 2003–2007. The ocean LHF fusion products estimated from ANN methods were 10–30 W m−2 lower than those of the four original ocean products (MERRA-2, JOFURO-3, ERA-I and GSSTF-3) and were more similar to observations.

Comparison of Performance of Predicting the Wear Amount of Tire Tread Depending on Sensing Information

“…The algorithms used to estimate tire wear amount were developed using the BPNN and the Automated Machine Learning (AutoML) method provided by PyCaret (version 2.3.10) [ 19 ]. First, the tire wear amount was predicted based on the features listed in Table 2 using AutoML, which is an ensemble method based on ten widely used ML algorithms, which include Linear Regression [ 20 ], Lasso Regression [ 21 ], Ridge Regression [ 22 ], Elastic Net Regression [ 23 ], Lasso Least Angle [ 24 ], Orthogonal Matching Pursuit [ 25 ], Bayesian Ridge [ 26 ], Passive Aggressive Regressor, Huber Regressor [ 27 ], and AdaBoost Regressor [ 28 ]. Most of the parameters for the ten ML methods provided by the AutoML function were kept as their default values (see Appendix A ).…”

ERTFM: An Effective Model to Fuse Chinese GF-1 and MODIS Reflectance Data for Terrestrial Latent Heat Flux Estimation

“…On the other hand, the ERTFM model is employed to merge satellite datasets and achieve accurate predictions and reconstructions and thus the accuracy of the machine learning method is highly dependent on the quality of the training data samples. Previous studies found that the larger the amount of training samples, the higher the accuracy estimations [57]. However, a large number of training datasets would lead to low computing efficiency, particularly for a large-scale area.…”

Can Indirect Evaluation Methods and Their Fusion Products Reduce Uncertainty in Actual Evapotranspiration Estimates?

“…The TCH method originated from Grubb's estimator (Grubbs, 1948), and then was further improved by Premoli and Tavella (1993). With the improvements, the TCH method can now be used for evaluating the uncertainty of more than three ET products (Chen et al., 2020; He, Xu, Bateni, et al., 2020; Long et al., 2014). The AA method reconciles the information from multiple ET products, using the AA of products as the reference against which to evaluate the performance of each product.…”