The CLD12 construction and its dependencies
Before CLD, one could attach a top-degree Chambert–Loir measure to a suitable metrized line bundle on a proper variety. A notation such as existed, but a local differential object was not supplied by that measure construction alone. CLD addresses this mismatch. The analytic space is allowed to be nonproper and need not be algebraic; the calculus must therefore be local and compatible with restriction.
In CLD12, Chapter 1 develops the real algebra of two copies of differential forms. Chapter 2 supplies polyhedral tropical images and the degree information needed to weight them. Chapter 3 uses these images as presentations of forms on the analytic space and proves integration, Stokes, and Green formulas. Chapter 4 takes continuous duals and proves the logarithmic divisor formula. Chapter 5 introduces positivity and constructs products for specified approximation classes. Chapter 6 connects local weights and residual line bundles with the existing model measures.
These references are to the chapters of CLD12, not to the numbering of this book. The positivity statements of CLD12 Chapter 5 use the test spaces constructed in its Chapters 3 and 4. The adjective positive refers to a cone of actual test forms. The equation is consequently more than a Hessian calculation on one fixed tropical chart. Similarly, the limit defining a mixed measure is an identity on all compactly supported tests, not merely a sequence of measures on one preferred skeleton.
The geometric and analytic parts meet in the canonical calibration. Algebraic or analytic multiplicities make a tropical integral independent of the moment map. This is what allows the distributional derivative to be intrinsic. The later delta-calculus retains exactly this requirement while enlarging the types of representatives that can be multiplied.
References. [CLD12] (Introduction and Chapters 1–6).
Conventions, potential classes, and sources
A current is a continuous linear functional on a space of test forms. Thus the expression has a mathematical meaning only after the test space, its topology, the positive cone, and the convention for degrees have been specified. A second question concerns the functions for which the expression is used. Positivity of a current, existence of smooth positive approximations, and numerical semipositivity of a model are different conditions. A comparison theorem must identify the map between the objects and state the hypotheses under which the conditions agree.
The book develops these comparisons in the following order. The real differential algebra and its cones are constructed in Section 7.1. Original analytic forms and their algebraic presentation are compared in Section 8.1; their continuous duals are defined in Section 9.1. The change from the 2012 to the 2025 definition of psh is treated in Section 11.2. Delta-forms and harmonic tropicalizations are then considered as separate constructions. The numerical and curvewise comparison results are stated with their actual regularity assumptions. The corrections in Section A.1 explain which literal statements from the sources cannot be used in this form.
Two conclusions organize the comparison. For continuous functions, the elementary arguments prove the chain with equality between and the continuous first decreasing-approximation stage; see Proposition 11.2.3, Proposition 11.2.4. For real PL functions on good Hausdorff boundaryless strictly analytic spaces over a nontrivially valued field, model semipositivity, membership in or , and classical curvewise psh are equivalent; see Proposition 19.2.11. This second comparison follows from model approximation and curve tests. Its proof does not assume the stronger general current criterion stated in CLD25.
Scope and conventions
The base field is complete and non-Archimedean. The original current calculus is described on a good Hausdorff analytic space of pure dimension , using test forms whose compact support avoids the analytic boundary. Statements about potentials and products are made on the boundaryless locus. The algebraic results of Gubler and Gubler–Künnemann will impose the additional assumptions that is algebraically closed and nontrivially valued and that comes from a separated algebraic variety. The harmonic-tropicalization results instead use good strictly analytic spaces over a nontrivially valued field. Each comparison retains these additional assumptions where they are needed.
An analytic boundary is the boundary of a Berkovich space or morphism. It is not a divisor deleted from an algebraic compactification. An algebraic open subvariety can have an algebraic boundary in a compactification and still have empty analytic boundary over .
We use additive weights No square of the norm is included. The symbol denotes the involution of the older Lagerberg convention; the conversion to is given in Proposition 7.1.6. Currents are indexed by bidegree unless lower indices are explicitly displayed. A measure on an -dimensional space has bidegree and bidimension .
Version-specific references
The reference labels below designate fixed texts. A theorem number in a later arXiv revision need not be its number in the published article. The dates in this table refer to the arXiv revisions actually used.
| Source | Version used | Principal comparison |
|---|---|---|
| CLD12 | 1204.6277v1; 27 April 2012; 124 pages | Original forms, currents, continuous current-psh functions, and a separate approximation class. |
| CLD25 | 1204.6277v2; 25 July 2025; 608 pages | New sign convention, generated psh sheaf, -forms, harmonic charts, and residual comparisons. |
| Gubler | 1303.7364v3; 5 October 2021 | Algebraic tropical charts for the original CLD sheaf. The underlying chapter was published in 2016. |
| GK17 | Published ANT article; 2017 | Delta-forms, delta-currents, and their intersection calculus. |
| GK19 | 1509.09079v2; September 2016 | Positivity, functorial psh conditions, and PL metric comparisons. |
| BGJK | 2003.08644v3; February 2021 | Real positive cones and counterexamples to strong-cone spanning and weak-positive order-zero claims. |
| GJR | 2111.05741v3; 7 March 2025 | Harmonic tropicalizations, weakly smooth forms, and strong currents. |
| GRPL | 2503.05513v1; 7 March 2025 | Semipositivity of real PL functions tested on analytic curves. |
| GRPSH | 2506.13548v2; 17 September 2025 | Classical psh defined by analytic curve tests after base extension. |
The 2025 CLD text is substantially different from its 2012 version. In particular, its psh definition is not obtained merely by renumbering the old definition. Conversely, a corrected statement in a later paper does not justify reading the earlier wording as if the correction were already present. The book therefore records both the mathematical definition and the source in which it occurs.
Remark 0.2.1 (Status of source comparisons). A theorem imported below is cited with a usable statement and its hypotheses. Appendix A uses three different descriptions: a false literal statement has a displayed counterexample; a typographical inconsistency has a local correction justified by the surrounding definitions; and a proof issue identifies a step not justified by the printed argument. A proof issue alone does not prove that the theorem is false. In particular, the deeper PL comparison stated in CLD25 is recorded separately from the statements that we disprove by elementary examples.
Index of notation
| Symbol | Meaning | Definition |
|---|---|---|
| Real superforms on an open subset of a real vector space | Definition 7.1.1 | |
| , | Older involution and CLD25 graded operator | Proposition 7.1.6 |
| Original smooth forms from analytic-unit tropicalizations | Definition 8.1.1 | |
| , | Ordinary and -smooth sheaves in CLD25 | Definition 17.3.2 |
| CLD25 forms represented by current-harmonic PL charts | Definition 17.3.4 | |
| Original test forms, with their presentation topology | Definition 9.1.1 | |
| Dual of original tests of degree | Definition 9.1.2 | |
| , | Currents obtained by integration against a function or form | Section 9.1 |
| Finite continuous functions with positive original curvature current | Definition 11.1.1 | |
| Local uniform limits of original smooth psh functions | Definition 11.2.1 | |
| , | CLD25 generated psh and continuous generated sheaves | Definition 11.2.2 |
| Ordinal stages of the decreasing-limit construction | Definition 11.2.2 | |
| Classical psh defined by analytic extension and curve tests | Definition 19.2.1 | |
| GJR weakly smooth forms from harmonic tropicalizations | Definition 17.2.5 | |
| Piecewise smooth forms with polyhedral derivatives | Theorem 17.2.6 | |
| Strong currents, dual to weakly smooth tests | Definition 17.2.10 | |
| Derivatives taken on individual polyhedral pieces | Section 17.2 | |
| , | Delta-forms and generalized delta-forms in the algebraic category | Definition 16.1.2, Definition 16.1.1 |
| Delta-currents, dual to compactly supported delta-forms | Definition 16.1.3 | |
| -PL | Piecewise linear functions with real coefficients, defined through local model data | Definition 19.1.4 |
| Curvature current of a metrized line bundle | Definition 11.1.6 |
Geometric and analytic notation used later
| Symbol | Meaning and normalization | Location |
|---|---|---|
| , with . | 5.2.1 | |
| Character lattice and its integral dual. | 5.1.2 | |
| , . | 5.1.2 | |
| ; the saturated tangent lattice. | 5.6.1 | |
| Initial degeneration, including scheme multiplicities. | 5.10 | |
| Gauss section of torus tropicalization. | 6.1.1 | |
| Completed residue field of an analytic point. | 1.6.2 | |
| Residual transcendence degree and relative rational value-group rank. | 3.8.1 | |
| in fixed Euclidean coordinates. | 9.13 | |
| for bidimension . | 9.13 | |
| Euclidean radial density, not generally an integral tropical weight. | 9.14.3 | |
| Integration current of a weighted tropical cycle. | 15.1.1 | |
| , ; boundary operators on delta-forms. | 16.5 | |
| Stable tropical intersection with lattice-index weights. | 15.3 | |
| ; Alexandrov normalization. | 21.27 | |
| for a convex real potential. | 21.5.1 | |
| ; support function. | 21.27 | |
| ; convex Legendre transform. | 21.26.1 | |
| Concave roof; sign conventions are fixed by the chosen metric weight. | 21.8 | |
| Normalized mixed metric energy relative to a fixed reference metric. | 20.11.1 | |
| Psh envelope in the stated global metric class. | 19.15 | |
| on a metric graph; positive Dirichlet sign. | 20.12.1 | |
| ; graph Dirichlet pairing. | 20.15.1 | |
| , . | 20.15.5 | |
| Trace weighted by local graph degrees. | 23.5.2 | |
| Multi-tangent coefficient space and its dual. | 7.8.2 | |
| Bott–Chern and Aeppli groups of the specified double complex. | 24.11.4 | |
| Projection to the constant Fourier mode on a real torus. | 24.11.1 |
The symbol can denote a polytope or an envelope operator, but never both in the same formula. Graph Laplacians have the sign ; Euclidean Laplacians in the distributional potential proof have the usual sign . These conventions serve different integration-by-parts formulas and must be retained when transferring a calculation.
Chapter concordance for CLD25
The CLD25 source architecture has a useful division of labour. Its Chapters 1–9 build a calculus on abstract tropical spaces. Chapters 10–16 prove that Berkovich analytic spaces supply the required structures and establish analytic current theorems. Chapters 17–19 apply the calculus to metrized bundles, residual intersection theory, higher-codimension formulas, and abelian varieties.
| Chapters | Objects constructed | Why they are needed |
|---|---|---|
| 1–3 | Superforms, paralinéaire spaces, G-tropical spaces and G-forms | A chart-independent language for differential expressions that may only be smooth on pieces. |
| 4–6 | Calibrations, integration, refined tropical structures and tropical spaces | Intrinsic integrals, compatible smooth forms, balancing, and Stokes. |
| 7 | Currents, positive and skeletal currents, harmonic charts | Duality, measure coefficients, structured singular currents, and extra coordinates. |
| 8 | Generated psh classes, CLN estimates, continuous products, dpsh, polar sets | A local pluripotential calculus with explicit approximation and independence conditions. |
| 9 | Products of spaces and currents, Gysin construction, tropical intersection | Intersections expressed through differential operations on products. |
| 10–11 | Graded reductions, residual intersection numbers, analytic geometry | Dimension, multiplicity, descent, and the residue geometry behind calibrations. |
| 12–14 | Analytic skeleta, richness, analytic degrees, canonical calibrations | Verification that the abstract tropical machinery applies intrinsically to analytic spaces. |
| 15–16 | Analytic forms and currents, finite maps, cycles, Poincaré–Lelong, support and polar results | The analytic realization of the first part and theorems involving analytic singular sets. |
| 17 | Metrized vector bundles, PL and formal metrics, residual bundles | Geometric sources of potentials and numerical positivity. |
| 18 | Curvature products, curve currents, residual formulas, Bochner–Martinelli and Levine | Explicit Chern and Green constructions, including higher codimension. |
| 19 | Cubist metrics, uniformization, positivity and canonical measures | A substantial application where harmonic coordinates and residual geometry interact. |
The abstract first part allows a differential argument to be proved once on a tropical space and then applied analytically. The price is that the hypotheses defining a tropical space are substantial. One cannot take an arbitrary weighted topological polyhedron, call it tropical, and inherit every theorem. Nor can one replace a skeletal current by an arbitrary current supported on a skeleton: the definition imposes compatibility with all charts.
Three changes especially affect reading. The psh sheaf is now generated from smooth positive functions by a closure procedure, so current positivity is a comparison statement. G-forms and harmonic forms are explicitly distinguished from ordinary smooth forms. Finally, the metric chapters use residual bundles and structured currents to obtain results beyond the original approximation-based measure theory. Each change creates new comparison questions, rather than identifying every old and new object by definition.
References. [CLD25] (Table of contents, Introduction, Chapters 1–19).