<p>The lower and upper approximations in rough set theory are important for dealing with uncertain knowledge. The covering rough set is an important part of the rough set theory and is suitable for dealing with numerical data. With the development of information technology, the information is constantly updated and changed in various information systems, and it is crucial to efficiently obtain the lower and upper approximations in dynamic environments. In this paper, the matrix-based methods for updating the approximate operators of local multigranulation neighborhood covering rough sets when objects are added or deleted are mainly investigated. Firstly, the local multigranulation neighborhood covering rough set model is introduced. Then, we analyze the relationships between the prior matrices and the updating matrices when objects change. Meanwhile, the dynamic updating processes for lower and upper approximations are proposed. The time complexity analyses of algorithms theoretically prove the efficiency of dynamic algorithms compared with static ones. To illustrate the effectiveness of the proposed dynamic algorithms, six datasets from the UCI are employed for comparative experiments.</p>