Introduction

A local differential theory on a Berkovich space must combine two kinds of data. Analytic functions supply logarithmic coordinates, while integral tangent lattices and multiplicities determine how those coordinates are integrated. The work of Chambert–Loir and Ducros constructs forms and currents from these data. Subsequent work develops delta-form intersections, harmonic coordinates, and comparisons with metrics and algebraic intersection theory. This book arranges these constructions so that their hypotheses and their precise points of contact remain visible.

The first basic issue is integration. Tropical multiplicities are not decorative weights: the change-of-variables formula contains a lattice index, and balancing cancels boundary contributions at internal faces. Chapters 1–6 establish the geometric background, and Chapter 8 explains the passage to an intrinsic analytic sheaf and its integrals. The Gauss-point and lattice calculations are proved directly. Polyhedrality, the general balancing theorem, faithful tropicalization and the analytic calibration theorems remain explicitly cited geometric inputs.

The next issue is positivity. In the real superform algebra, the three positivity cones need not agree in middle degree. The ordinary positive cone controls all symmetric coefficients; the strong cone need not span them. Chapters 7–9 develop this distinction, prove the applicable order-zero estimates, and construct continuous scalar multiplication by measures attached to actual tests. The intrinsic proof is made in an explicit algebraic CLD setting. It does not silently identify every analytic or harmonic quotient with that setting.

For potential theory, the construction d′d′′(uT)d'd''(uT) separates multiplication from differentiation. Positivity of this current requires an admitted class of potentials, and the available limit theorem depends on that class. Chapter 12 proves the normalized regularized maximum and its local gluing properties. Chapter 13 proves mixed-product continuity and symmetry for the explicit smooth-approximation class. Chapter 14 explains why independence of an approximating sequence does not imply independence after changing the potential by an arbitrary current-harmonic function.

The enlarged calculi address different needs. Delta-forms retain corner terms as objects that can participate in intersection products. Weakly smooth forms permit harmonic coordinates which need not be original smooth coordinates. Model and curvewise criteria compare numerical semipositivity with analytic positivity in specified regularity classes. Chapters 15–21 keep these operations separate until a comparison theorem supplies a map between them. The discussion of Green objects and descent then records the additional requirements on representatives, tests and support.

The two CLD manuscripts are treated as different versions of a developing theory. Corrections are attached to exact source statements. A false real-algebra claim, an unsupported analytic proof step, and a typographical formula error have different consequences. The source concordance makes these distinctions explicit. The present text is a working reference manuscript, with full proofs where supplied and identified external inputs elsewhere; it is not a claim that every gap in the literature has been resolved.

The longer-term purpose is to turn this reference manuscript into a proof-complete, searchable foundation for non-Archimedean geometry. An imported result remains an imported result until its proof and the proofs of its non-elementary prerequisites have been incorporated and checked. The proof architecture explains this standard and identifies the first foundational dependencies to be closed.