In general, the component-wise partial derivative of a tensor field is not again a tensor field, since it doesn't has the correct transformation law. This is one of the reasons why one introduces covariant derivatives at the first place. Now, I stumbled over the following observation:
Let $M$ be a smooth manifold and $\omega_{\mu}dx^{\mu}\in\Omega^{1}(M)$ a $1$-form. Then, it is not too hard to see that
$$(\mathcal{L}_{\partial_{\nu}}\omega)_{\mu}=\partial_{\nu}\omega_{\mu}$$
for some fixed index $\nu$. This formula can be obtained by using for example Cartan's formula to calculate. In particular, since $\mathcal{L}_{X}$ maps a $1$-form to a $1$-form, we see that $\partial_{\nu}\omega:=(\partial_{\nu}\omega_{\mu})d x^{\mu}$ is a well-defined $1$-form. Similar calculations can be done for arbitrary $k$-forms. So, does this imply that partial derivative of forms are in fact again forms?
In a different frame the vector field $D\equiv \partial_\nu= \delta^\mu_\nu\partial_\mu $ will no longer have constant components $\delta^\mu_\nu$. You therefore need to include their transformation when you transform $\partial_\nu \omega_\mu$. This will bring you to the usual $({\mathcal L}_{D}\omega)_\mu = D^\nu\partial_\nu\omega_\mu + \omega_\nu \partial_\mu D^\nu $.