Reproducing the Madden‐Julian Oscillation (MJO) remains a challenge in many climate models. This study compares the MJO simulation skills in 17 pairs of Coupled Model Intercomparison Project Phase 5 (CMIP5) and Coupled Model Intercomparison Project Phase 6 (CMIP6) simulations. Generally, the MJO characteristics are better reproduced in CMIP6, their inter‐model spreads are obviously reduced in CMIP6. However, most CMIP6 models still underestimate the frequency of MJO initiation over the Indian Ocean and the MJO amplitude, exaggerate the barrier effect of the Maritime Continent (MC) on MJO propagation. Models with lower MJO simulation skills in their CMIP5 versions show improvements in CMIP6, while other models tend to show degraded MJO simulation skills in CMIP6. The possible reasons for the apparent improvements of MJO simulations from CMIP5 to CMIP6 in six models was further explored. It was found that the enhanced land precipitation over the MC prior to the MJO convection entering this region is more obvious in CMIP6.

The cool skin effect, known as the temperature difference (ΔT) across the skin layer of sea surface, is of vital importance for the accurate computation of the latent heat flux (LHF). The observed features of ΔT in the South China Sea are analyzed using in situ data from a buoy platform over an ∼6‐week period. Only nighttime data are used to exclude the possible warm layer effect. The positive values of ΔT falling into the range of 0–1 K comprise 95% of the data, and the most frequently observed values occur in the range of 0.4–0.6 K (38%). The cool skin model in the Coupled Ocean‐Atmosphere Response Experiment 3.0 algorithm is then validated against those observations. The cool skin model has an efficient but insufficient ability to reduce the overestimation of the LHF. The overestimation of the LHF is reduced to 9.5% from 18.0%, leaving nearly half of the biases in the LHF unresolved. The Saunders constant (λ) in the cool skin model is markedly underestimated, leading to a much weaker prediction of ΔT. A strong linear relationship exists between the mean values of λ and the LHF with a slope of −0.9 W m⁻². With an approximately doubled λ, the biases in ΔT and in the LHF could be eliminated. Considering the possible uncertainties in sensors, the value of λ is estimated as 11.6 ± 6.7 in the current study.

Simulations of the global oceanic latent heat flux (LHF) in the CMIP5 multimodel ensemble (MME) were evaluated in comparison with 11 LHF products. The results show that the mean state of LHF in the MME coincides well with that in the observations, except for a slight overestimation in the tropical regions. The reproduction of the seasonal cycle of LHF in the MME is in good agreement with that in the observations. However, biases are relatively obvious in the coastal regions. A prominent upward trend in global-mean LHF is confirmed with all of the LHF products during the period of 1979–2005. Despite the consistent increase of LHF in CMIP5 models, the rates of increase are much weaker than those in the observations, with an average of approximately one-ninth that in the observations. The findings show that the rate of increase of near-surface specific humidity qa in MME is nearly 6 times that in the observations, while the rate of increase of the near-surface wind speed U is less than one-half that in the observations. The faster increase of qa and the slower increase of U could both suppress evaporation, and thus latent heat released by the ocean, which may be one of the reasons that the upward trend of LHF in the MME is nearly one order of magnitude lower than that in the observations.

The China-France Oceanography SATellite (CFOSAT), launched on 29 October 2018, is a joint mission developed by China and France. To evaluate the CFOSAT wind product, L2B swath data with a spatial resolution of 25 × 25 km were compared with in situ measurements between December 2018 and December 2020. The in situ measurements were collected from 217 buoys. All buoy winds were adjusted to 10 m height using a simple logarithmic correction method. The temporal and spatial separations between the CFOSAT and in situ measurements were restricted to less than 30 min and 0.25°. The results indicate that the CFOSAT wind retrievals agree well with the buoy measurements. The root mean square errors (RMSEs) of wind vectors were 1.39 m s⁻¹ and 34.32° and negligible biases were found. In the near shore under rain-free conditions, the RMSEs were enhanced to 1.42 m s⁻¹ and 33.43°. Similarly, the RMSEs were reduced to 1.16 m s⁻¹ and 30.41° offshore after the rain effect was removed. After winds less than 4 m s⁻¹ were removed, the RMSE of wind directions was reduced to 19.69°. The effects of significant wave height, air-sea temperature difference, sea surface temperature, atmospheric pressure and ocean surface current on the wind residuals were assessed. The performance of wind retrievals under the passage of tropical cyclones was evaluated. The evaluation results show that the CFOSAT wind retrievals satisfy the accuracy requirements of scientific research, although some improvements are needed to enhance the performance.

Five latent heat flux (LHF) products are evaluated based on in situ observations in the South China Sea (SCS), including the ECWMF ERA‐Interim (ERA‐I), the NCEP2, the Objectively Analyzed air‐sea Fluxes (OAFlux), the Japanese 55 year Reanalysis (JRA55), and the TropFlux data sets. The results show that there are good correlations between the LHF products and observations, ranging from 0.68 to 0.74. However, mean biases of −8 to 40 W m⁻² exist in the LHF products with respect to the observations. For root‐mean‐square errors, the OAFlux data set is the closest to the observations, followed by ERA‐I and TropFlux, while the NCEP2 data set shows significant overestimation. It is found that the biases in the near‐surface‐specific humidity are most correlated with the biases in the LHF products, followed by the biases in the near‐surface wind speed, air temperature, and sea surface temperature. The biases in the LHF products have a prominent seasonal variation that is 25 W m⁻² higher in boreal winter than in summer. Using the thermal equation, it is shown that the tendency errors of the mixed‐layer temperature estimated by the biases in the LHF products vary from −2.0 to 3.5°C/month in the SCS. When all of the products are averaged, the errors are reduced to a range of −0.7 to 1.5°C/month. It is noteworthy that the errors in summer are more obvious than those in winter, since a thinner mixed layer in the summer can amplify the effect of even a small bias in the LHF.