SST biases over the Northwest Pacific and possible causes in CMIP5 models

In this paper, the features and possible causes of sea surface temperature(SST) biases over the Northwest Pacific are investigated based on a mixed-layer heat budget analysis in 21 coupled general circulation models(CGCMs) from phase 5 of the Coupled Model Inter-comparison Project(CMIP5). Most CMIP5 models show cold SST biases throughout the year over the Northwest Pacific. The largest biases appear during summer, and the smallest biases occur during winter. These cold SST biases are seen at the basin scale and are mainly located in the inner region of the low and mid-latitudes. According to the mixed-layer heat budget analysis, overestimation of upward net sea surface heat fluxes associated with atmospheric processes are primarily responsible for the cold SST biases. Among the different components of surface heat fluxes, overestimated upward latent heat fluxes induced by the excessively strong surface winds contribute the most to the cold SST biases during the spring, autumn, and winter seasons. Conversely, during the summer, overestimated upward latent heat fluxes and underestimated downward solar radiations at the sea surface are equally important. Further analysis suggests that the overly strong surface winds over the Northwest Pacific during winter and spring are associated with excessive precipitation over the Maritime Continent region,whereas those occurring during summer and autumn are associated with the excessive northward extension of the intertropical convergence zone(ITCZ). The excessive precipitation over the Maritime Continent region and the biases in the simulated ITCZ induce anomalous northeasterlies, which are in favor of enhancing low-level winds over the North Pacific. The enhanced surface wind increases the sea surface evaporation, which contributes to the excessive upward latent heat fluxes. Thus, the SST over the Northwest Pacific cools.

ROBUST DESIGN MODELS FOR CUSTOMER-SPECIFIED BOUNDS ON PROCESS PARAMETERS

Robust design (RD) has received much attention from researchers and practitioners for years, and a number of methodologies have been studied in the research community. The majority of existing RD models focus on the minimum variability with a zero bias. However, it is often the case that the customer may specify upper bounds on one of the two process parameters (i.e., the process mean and variance). In this situation, the existing RD models may not work efficiently in incorporating the customer’s needs. To this end, we propose two simple RD models using the ε?constraint feasible region method - one with an upper bound of process bias specified and the other with an upper bound on process variability specified. We then conduct a case study to analyze the effects of upper bounds on each of the process parameters in terms of optimal operating conditions and mean squared error.