arXiv2026
Drinfeld modules, introduced by D.~V.~Drinfeld in the 1970s, were originally developed as a function field analogue of elliptic curves and have since become a central tool in the Langlands program over function fields. The theory has been highly developed and shares deep connections with many areas, including algebraic geometry, number theory, and coding theory. Despite these advances, the explicit construction of Drinfeld modules over non-polynomial coordinate rings remains a largely open problem. Indeed, aside from the classical polynomial case $\mathbb{F}_q[t]$, explicit formulas for Drinfeld modules are known only in very limited situations. Let \(E\) be an elliptic curve over a finite field \(\mathbb{F}_q\) with a fixed rational point \(\infty\), and let \(\mathbf{A} = H^0(E\setminus\{\infty\}, \mathcal{O}_E)\) be its coordinate ring. Rank-one Drinfeld \( \mathbf{A} \)-modules were explicitly constructed by Green and Papanikolas, providing the first systematic example beyond the polynomial case. However, the rank-two case has remained completely inaccessible until now. This paper solves the rank-two case of this open problem in a fully explicit manner. Precisely, we develop an explicit theory for rank-two sign-normalized \(\mathbf{A}\)-Drinfeld modules over an algebraically closed \(\mathbf{A}\)-field \(L\), building upon the rank-one framework. We determine the structure of the associated Anderson motive \(M_ϕ\) and prove that it is generated by three elements subject to a single quadratic \(τ\)-relation, which we derive in closed form using the geometric parameters of the underlying elliptic curve.As a consequence, we find that the moduli space of sign-normalized rank-two Drinfeld \(\mathbf{A}\)-modules is an open domain $Y\neq 0 $ inside a supersingular curve \(Y^{q+1} = π(X) \), where \(π\) is an explicit polynomial of degree \(2q+1\).