The main derivative operator in GA is the vector derivative $\nabla = {{\mathbf{e}}^j}{\partial _j}$, with $\{ {{\mathbf{e}}^j}\}$ the reciprocal base with respect to an arbitrary frame field $\left\{ {{{\mathbf{e}}_j}} \right\}$.
Vector Derivative $\nabla$
- yields the gradient when operating on a scalar field $\phi =
\phi (x)$
\[ \nabla \phi (x) = {{\mathbf{e}}^j}{\,\partial _j}\,\phi (x) \]
- yields the geometric product when operating on a vector field
\[ \nabla {\mathbf{a}} = \nabla \cdot {\mathbf{a}} + \nabla \wedge {\mathbf{a}} \]
- yields the divergence through the inner product
\[\nabla \cdot {\mathbf{a}} = {{\mathbf{e}}^j} \cdot {\partial _j}{\,\mathbf{a}} = {\partial _j}{\,a^j}{\qquad}{a^j}: = {{\mathbf{e}}^j} \cdot {\mathbf{a}}\]
- yields the curl through the wedge product
\[ \nabla \wedge {\mathbf{a}} = {{\mathbf{e}}^i} \wedge {\,\partial _i}{\,\mathbf{a}}= {\partial _i}{\, a^j}{{\,\mathbf{e}}^i} \wedge {{\mathbf{e}}_j} \]
The operator $\nabla $ has all the algebraic properties of a vector and can act over any multivector with the inner, outer or geometric product. For example, if $\phi $ is a scalar field and ${\mathbf{a}}$ and ${\mathbf{b}}$ two vector fields, then