Before we get into what a Merkle root in blocks of blockchain is, let's understand the structure of blockchain. A block is made up of two parts; the first part is the block header and the second part is the set of transactions of that block. The block header contains information such as the previous block hash (it's actually a hash of the previous block's header), timestamp, Merkle root, and information related to achieving consensus.
At the time of sync, while downloading a block a node downloads the block header and the block's transactions. Now, how would the receiving node know that these transactions are actually part of that block and are in the correct order? Every block is identified by a unique hash, but the block hash is not part of the block header and is uniquely calculated by every node after downloading the block; therefore we cannot use the idea of a block hash. Instead, we can rely on something like a transactions hash; a hash stored in the block header...