We consider the optimal control problem governed by diffusion convection reaction equation without control constraints. Proper orthogonal decomposition (POD) method is used to reduce the dimension of the problem. The POD basis which is generated via the snapshot ensemble depending on a set of parameters may not give accurate results when it is used to approximate the problem depending on a different set of parameters [1]. For an accurate reduced solution of the perturbed problem, the POD basis must be regenerated for each parameter, which is costly. In this paper, motivated by the study [2], we include the sensitivity information to enrich the low-dimensional subspace by extrapolating and expanding the POD basis in case of perturbation of a parameter in the convection term. We construct the snapshot matrix using the state solution, the adjoint solution or a combination of them [3]. We compare these different bases in terms of accuracy and complexity by measuring the error and the computational time, and investigate the advantages and main drawbacks of them.