You are an expert differential geometer, mathemati...

You are an expert differential geometer, mathematical expositor, and LaTeX editor. You are given several overlapping drafts of lecture notes on Semi-Riemannian geometry, with particular emphasis on the Riemannian and Lorentzian cases.

Your task is not merely to edit, concatenate, or summarize these drafts. Treat them as raw source material and transform them into a single, coherent, mathematically rigorous, pedagogically exceptional, publication-ready course handout.

1. Final deliverable

Produce ONE complete, self-contained, compilable .tex file.

The result should read as a polished textbook-style set of lecture notes rather than as a collection of revised drafts. It should have:

a clear title and subtitle;
a short preface explaining the goals and philosophy of the notes;
an explicit list of prerequisites;
a notation-and-conventions section;
a logically designed table of contents;
coherent chapters/sections/subsections;
definitions, propositions, theorems, remarks, warnings, worked examples, and exercises;
professional TikZ figures;
an annotated bibliography;
consistent notation, terminology, theorem numbering, cross-references, and formatting throughout.

The .tex source must compile as a standalone document without requiring manual repair.

Do not output commentary about what you changed. Output the finished LaTeX document itself.

2. Intended audience

Write for mathematically mature students who may know basic multivariable calculus, linear algebra, and elementary smooth-manifold terminology, but who have:

no prior knowledge of Semi-Riemannian geometry;
no prior knowledge of Lorentzian geometry;
no prior knowledge of affine/Levi-Civita connections;
no prior knowledge of covariant differentiation;
no prior knowledge of curvature.

Consequently, no central concept concerning connections or curvature may be treated as “standard” and skipped.

Every important new definition should be motivated geometrically before or immediately after it is formalized.

Use the principle:

Geometry first; formulas second; abstraction only when it clarifies the geometry.

Avoid unnecessary generality. In particular, develop connections and covariant derivatives on the tangent bundle $TM$ , not on arbitrary vector bundles unless a very brief remark is pedagogically indispensable.

3. Editorial treatment of the supplied drafts

Use all supplied drafts critically.

Do NOT simply choose one draft as the base document or concatenate passages from different drafts.

Instead:

Identify the strongest explanations, examples, computations, exercises, and figures in each draft.
Merge overlapping treatments into one coherent exposition.
Remove repetition and inconsistencies.
Reorder material whenever this improves the pedagogical progression.
Standardize notation and terminology.
Detect and correct mathematical mistakes, ambiguous statements, misleading simplifications, and inconsistent sign conventions.
Supply missing intermediate arguments when the drafts move too quickly.
Add standard material needed to make the resulting text genuinely standalone.
Preserve particularly effective geometric explanations from the drafts whenever possible, but rewrite them so that the final document has one consistent authorial voice.

The finished text must feel as though it was written from the beginning as one book.

4. Mathematical conventions

Choose and state all conventions explicitly near the beginning of the document and adhere to them everywhere.

For Lorentzian geometry, preferably use signature

(-,+,\ldots,+),

unless there is an overwhelming reason in the drafts to do otherwise.

Explicitly state the chosen conventions for:

the Lorentzian signature;
the Riemann curvature tensor $R(X,Y)Z$ ;
the $(0,4)$ -curvature tensor;
sectional curvature;
Ricci curvature;
scalar curvature;
the second fundamental form, if used.

Whenever a formula is convention-dependent, warn the reader.

A recurring pedagogical theme should be:

Which facts survive unchanged from Riemannian geometry, which facts require sign modifications, and which familiar positive-definite intuitions genuinely fail in Lorentzian geometry?

5. Desired mathematical arc

Design the course so that ideas arise naturally from geometric problems rather than appearing as disconnected formalism.

A possible progression is:

Part I. From metrics to Semi-Riemannian geometry

tangent vectors and tangent spaces as velocities;
vector fields;
symmetric bilinear forms;
nondegeneracy, index, and signature;
Sylvester's law of inertia;
Riemannian, Semi-Riemannian, and Lorentzian metrics;
Minkowski space as the fundamental Lorentzian model;
spacelike, timelike, null, and causal vectors;
the light cone;
time orientation;
comparison between Euclidean and Minkowskian geometry;
basic model spaces.

Part II. Why ordinary differentiation of vector fields fails

Build the connection from the geometric problem:

If $X(p)\in T_pM$ and $X(q)\in T_qM$ , these vectors live in different vector spaces. What could it mean to differentiate $X$ ?

Use this problem to motivate the connection rather than presenting the definition abruptly.

Develop carefully:

directional differentiation in Euclidean space;
why it does not immediately globalize;
affine connections on $TM$ ;
the covariant derivative $\nabla_XY$ ;
coordinate vector fields;
Christoffel symbols;
transformation issues and why Christoffel symbols are not tensor components;
torsion;
metric compatibility;
the Levi-Civita connection;
geometric meaning of the fundamental theorem of Riemannian/Semi-Riemannian geometry;
the Koszul formula;
systematic computations of Levi-Civita connections.

Emphasize what a connection does geometrically, not merely its axioms.

6. Vector fields along curves

Give this topic substantial attention.

Carefully distinguish:

a vector field $X$ defined on an open subset of $M$ ;
a vector field $V(t)\in T_{\gamma(t)}M$ along a curve $\gamma$ ;
the velocity field $\dot\gamma$ ;
the covariant derivative $D V/dt$ .

Explain why $dV/dt$ is generally meaningless intrinsically, because the vectors $V(t)$ belong to different tangent spaces.

Derive and interpret the coordinate formula

\frac{DV}{dt} = \left( \frac{dV^k}{dt} + \Gamma^k_{ij}\dot\gamma^iV^j \right)\partial_k.

Do not merely display this formula: explain geometrically what each term is correcting.

Provide several fully worked computations.

7. Parallel transport as a central geometric idea

Make parallel transport one of the conceptual centers of the notes.

Develop:

parallel vector fields along curves;
the equation $DV/dt=0$ ;
existence and uniqueness as an ODE;
parallel transport $P_\gamma$ ;
dependence on the path;
metric preservation by Levi-Civita parallel transport;
parallel frames;
transport around closed loops;
comparison of Euclidean, spherical, hyperbolic, and Lorentzian examples.

The reader should acquire a vivid geometric understanding of what “keeping a vector parallel” means when there is no ambient vector space in which tangent vectors can simply be translated.

Include explicit, nontrivial computations of parallel transport, for example on some of:

$\mathbb R^n$ ;
the Euclidean plane in polar coordinates;
a cylinder;
the round sphere $S^2$ ;
the hyperbolic plane;
Minkowski space in Cartesian and non-Cartesian coordinates;
simple warped-product Lorentzian metrics.

Where useful, accompany these examples with clear TikZ figures.

8. Holonomy

Introduce holonomy directly from parallel transport around loops.

Explain geometrically:

why returning to the same point does not necessarily return a vector to itself;
the holonomy group at a point;
dependence on loops;
how holonomy measures the failure of path-independent parallel transport;
the relationship between flatness and local path independence;
how small-loop holonomy anticipates curvature.

At least one example should explicitly calculate or geometrically demonstrate nontrivial holonomy.

The sphere is particularly valuable: parallel transport a tangent vector around a geodesic triangle and explain the resulting rotation.

Use this example to prepare the transition from parallel transport to curvature.

9. Geodesics

Develop geodesics from the connection rather than defining them merely by a coordinate ODE.

Explain:

\frac{D\dot\gamma}{dt}=0

as the intrinsic statement that a curve transports its own velocity parallel to itself.

Discuss:

affine parametrization;
the coordinate geodesic equation;
Christoffel-symbol calculations;
geodesics in Euclidean space;
great circles;
hyperbolic examples;
Minkowski geodesics;
causal character of Lorentzian geodesics.

Clearly distinguish the Riemannian length-minimizing intuition from the Lorentzian situation, where timelike geodesics locally maximize proper time under suitable conditions.

10. Curvature: make the geometry primary

Do not introduce curvature as an unexplained four-index object.

Build toward it through geometric questions:

Does parallel transport depend on the path?
What happens when one parallel-transports around a tiny loop?
Do covariant derivatives commute?
How does the geometry deviate from its flat tangent-space model?

Then introduce

R(X,Y)Z = \nabla_X\nabla_YZ - \nabla_Y\nabla_XZ - \nabla_{[X,Y]}Z.

Explain each term and why the Lie-bracket correction is necessary.

Give geometric interpretations before pursuing tensor symmetries.

Develop carefully:

curvature operator;
Riemann curvature tensor;
tensoriality;
basic symmetries;
Bianchi identity, with geometric commentary;
sectional curvature;
Ricci curvature;
scalar curvature.

For each curvature notion, answer explicitly:

What geometric information does this quantity measure?

In particular:

Sectional curvature

Explain it as curvature associated with a two-dimensional tangent plane and connect it to Gaussian curvature and the geometry of geodesics.

Ricci curvature

Explain it as a trace/average of sectional curvatures through a given direction, and discuss why it plays a distinguished role in Lorentzian geometry and general relativity.

Scalar curvature

Explain it as a further averaged trace, while making clear what geometric information is lost in passing from the full curvature tensor to Ricci and then to scalar curvature.

11. Curvature and infinitesimal parallel transport

Give special emphasis to the relation

curvature = infinitesimal holonomy.

Explain carefully how the curvature tensor measures the first nontrivial change produced by transporting a vector around a sufficiently small loop.

Include a schematic TikZ diagram of a small parallelogram or geodesic rectangle, showing the vector before and after parallel transport.

This should be one of the main conceptual payoffs of the course.

12. Riemannian versus Lorentzian geometry

Throughout the text, deliberately compare the two settings.

Use dedicated remarks or boxes such as:

“Riemannian intuition”
“Lorentzian warning”
“Same formula, different geometry”
“What fails in indefinite signature?”

Important issues include:

vectors of zero norm need not be zero;
orthogonal complements behave differently;
null vectors are orthogonal to themselves;
the unit “sphere” becomes hyperbolic/noncompact;
reverse Cauchy–Schwarz inequalities for timelike vectors;
causal cones;
proper time;
maximizing versus minimizing geodesics;
spacelike/timelike/null sectional planes where appropriate;
degeneracy problems for null planes.

Do not hide the Lorentzian geometry behind Riemannian notation. The indefinite-signature phenomena should be visible throughout.

13. Worked examples and computations

The document must contain many substantial worked examples.

Avoid examples that simply substitute numbers into a formula.

Prefer examples that reveal a genuine geometric phenomenon.

For each major topic, include examples of increasing sophistication:

a transparent model computation;
a coordinate computation;
a nontrivial geometric application.

Whenever Christoffel symbols are computed, show enough intermediate steps that a beginner can reproduce the calculation.

Whenever curvature is computed, make clear:

what metric is being used;
what coordinates are being used;
which Christoffel symbols vanish;
which terms survive;
what the final result means geometrically.

Useful recurring models include:

Euclidean space;
polar coordinates;
the round sphere;
cylinders and surfaces of revolution;
hyperbolic space;
Minkowski space;
Rindler-type coordinates;
de Sitter and anti-de Sitter space where appropriate.

Do not include a difficult calculation merely for its own sake. Every long computation should illuminate an idea.

14. TikZ figures

Create clean, professional TikZ illustrations whenever a picture materially improves understanding.

The diagrams should be mathematically meaningful, not decorative.

Useful figures include:

tangent spaces attached to different points of a manifold;
why vectors at different points cannot simply be subtracted;
vector fields along curves;
parallel transport;
parallel transport around a loop;
holonomy on the sphere;
a small loop illustrating curvature;
geodesics on a sphere;
light cones;
timelike, spacelike, and null directions;
causal cones along a Lorentzian manifold;
comparison of Euclidean circles and Minkowski hyperbolae.

Figures must:

compile using standard TikZ/PGF libraries;
use consistent styling;
have readable labels;
avoid unnecessary visual clutter;
have informative captions;
be referenced from the surrounding text;
remain understandable in print.

15. Pedagogical style

The exposition should be rigorous but conversational and explanatory.

Whenever possible, use the pattern:

Geometric question → motivating example → definition → computation → interpretation → warning/remark → exercise.

Important definitions should normally be followed immediately by:

a plain-language interpretation;
a simple example;
a nontrivial example or consequence.

Avoid phrases such as “clearly”, “obviously”, or “it is easy to see” at points where a beginner may genuinely need an argument.

Do not suppress intermediate computations merely to shorten the text.

At strategic points include concise “What to remember” boxes that summarize the conceptual content rather than merely listing formulas.

16. Exercises

Include carefully selected exercises throughout the course, not merely at the end.

Organize them by difficulty, for example:

★ foundational;
★★ intermediate;
★★★ challenging.

The exercises should be designed to reveal important ideas rather than merely test algebraic manipulation.

Include exercises involving:

interpretation of Christoffel symbols;
computing Levi-Civita connections;
differentiating vector fields along curves;
constructing parallel vector fields;
explicit parallel transport;
geodesic equations;
holonomy;
calculating curvature;
constant-curvature spaces;
comparison of Riemannian and Lorentzian phenomena;
causal geometry in Minkowski space;
examples where Riemannian intuition fails.

Include a substantial selection of complete solutions or detailed solution sketches, depending on the scale of the final document.

A good exercise should ideally teach the student something that was not completely explicit in the preceding exposition.

17. Mathematical rigor

The document is intended for serious students.

Therefore:

distinguish definitions from consequences;
state hypotheses precisely;
never confuse local and global statements;
distinguish coordinate-dependent objects from tensors;
distinguish affine from metric notions;
explain where nondegeneracy is used;
distinguish the Riemannian and Lorentzian cases whenever positivity matters;
do not use curvature identities before proving or stating them;
ensure that indices and signs are consistent;
verify all displayed calculations.

Do not propagate an assertion from a source draft merely because it appears authoritative.

18. Bibliography

Finish with an annotated bibliography of particularly pedagogical references on:

Riemannian geometry;
Semi-Riemannian geometry;
Lorentzian geometry;
differential geometry and connections;
global Lorentzian geometry / causality where appropriate.

For every recommended book, give a short annotation explaining:

its level;
its pedagogical strengths;
which parts are especially useful;
whether it is best suited for a first reading, a second course, exercises, geometric intuition, or reference.

Prefer genuinely influential and pedagogically strong texts.

Do not invent bibliographic information. If publication details cannot be stated reliably, use a conservative citation rather than fabricating an edition, year, or publisher.

19. LaTeX and production quality

Use professional LaTeX practices.

The source should include, as appropriate:

amsmath, amssymb, amsthm, mathtools;
geometry;
microtype;
enumitem;
booktabs;
xcolor;
tikz with explicitly declared libraries;
hyperref;
cleveref if helpful;
a well-designed theorem environment system.

Define macros for frequently used mathematical notation rather than repeating fragile code.

Do not introduce unnecessary packages.

Avoid compilation warnings where reasonably possible.

Check:

all environments are closed;
all references resolve;
all equation labels are unique;
all figures are referenced;
no undefined commands remain;
TikZ code compiles;
bibliography references are internally consistent.

20. Final quality test

Before returning the document, silently audit it as both:

an expert differential geometer checking correctness, and
a student encountering connections and curvature for the first time.

Ask:

Can the student understand why a connection is needed before seeing its axioms?
Can the student explain covariant differentiation geometrically?
Can the student distinguish a vector field on $M$ from a vector field along a curve?
Can the student explain parallel transport without relying only on formulas?
Can the student explain holonomy geometrically?
Can the student explain curvature as infinitesimal failure of parallel transport to be path-independent?
Can the student say what sectional, Ricci, and scalar curvature each measure?
Can the student identify precisely where Lorentzian geometry differs from Riemannian geometry?
Are there enough fully worked examples to make the formalism usable?
Are the diagrams genuinely explanatory?
Does the whole document have a coherent narrative arc?

If the answer to any of these is no, revise before producing the final output.

The ultimate goal is a handout that a strong student with no previous knowledge of connections, curvature, or Lorentzian geometry could read independently and emerge with both computational competence and genuine geometric intuition.