Dimensionality reduction techniques on scientific data are currently a focused approach to understand the underlying scientific knowledge in a dataset resulted from scientific experiments and simulations. The sole purpose of the survey paper is to provide comprehension of different dimensionality reduction techniques which are used in the field of scientific data mining. Feature extraction and feature selection are the important techniques of dimensionality reduction; the former removes certain features by way of transformation, where as the later reconstructs its features into a lower dimension space without impairing its initial characteristics. This paper presents various techniques used for mining the scientific data.