The authors review the major progresses in the targeting observations and field campaigns for Tropical Cyclones （TCs） using their proposed nonlinear optimal perturbation methods. They used condition nonlinear optimal perturbation， which aims at reducing the initial errors for numerical models， to identify the sensitive regions for TC track and intensity forecasts. This guides both FY-4 satellite and dropsonde in the field campaigns to effectively collect valuable observation data. Furthermore， they investigated the sensitivity of TC intensity forecast to model errors using nonlinear forcing singular vector， which indicates both the sensitive regions and variables in troposphere and the sensitive regions of sea surface temperature. Simultaneously， they also identified the sensitive regions and variables in the boundary layer during a rapid intensification process of a TC using ensemble perturbation method. In final the unresolved problems， possible solutions， and future recommendations in the targeting observations for TCs are discussed.