Bash Variables – Difference Between Environment Variable and Shell Variable

In the unix model, starting another program involves two primivites:

• fork() creates an (almost) identical copy of the calling process. The new process is called the child process and the original process is called the parent process. The child process runs the same code as the original, has the same permissions, has the same environment, and receives a copy of the mutable data memory of the parent process. The most visible difference between the two processes is that they have different process IDs and different parent process IDs (the child's PPID is the parent's PID).
• execve() replaces the code and data of the current process by code and data loaded from an executable file. This system call takes the new environment of the process as an argument.

Most high-level functions built around fork() and execve() pass the process's current environment to execve(). Thus, unless the process changes its own environment or calls execve() directly, the called program will inherit the calling program's environment.

Shells normally pass their environment to the programs they call. You can change the environment of a shell at any time by assigning a value to an environment variable (foo="some value"; you must call export foo if the variable isn't in the environment already), or remove a variable from the environment by unsetting it (unset foo). If you want to start an external program with different or additional environment variables, there is a shortcut syntax:

foo="some value" mycommand

is roughly equivalent to

(foo="some value"; export foo; exec mycommand)

(where the parentheses restrict the scope of the setting of foo).