Abstract: This paper introduces the def i nition of conditional nonlinear optimal perturbation (CNOP), and the applications of the CNOP in atmosphere and ocean studies. The CNOP approach is expanded as that related to initial perturbation (CNOP-I), related to parameter perturbation (CNOP-P), and the combined both of CNOP-I and CNOP-P, according to the different perturbation types. The CNOP-I approach has been applied to the predictability studies of ENSO events, Kuroshio path anomalies, blocking, nonlinear stabilities of thermohaline circulation and grassland ecosystem. The CNOP-I has been further employed to explore the target observation of typhoon. The sensitive region could be identif i ed by using the CNOP-I approach. The forecast skill may be improved by adding more adaptive observations in the sensitive region. The CNOP-P approach has been applied also to Kuroshio path anomalies, nonlinear stabilities of thermohaline circulation and grassland ecosystem. Here, we carried out a numerical simulation how to obtain the CNOP with the Burgers equation through building the tangent linear model and adjoint model. The result shows that the CNOP can be calculated by using the Burgers equation, the tangent linear model and the adjoint model with nonlinear optimization algorithm; It supplies a guide to a beginner to learn the CNOP and a reference for employing the CNOP to other applicable subjects.