RockFall Toolkit: This toolkit provides two linked rockfall trajectory simulators: a 2D slope-profile model for quick screening along a single representative cross-section, and a 3D DEM-based raster model for spatial hazard mapping across real terrain. Both implement a stochastic, process-based lumped-mass/rigid-body trajectory model — releasing many randomized blocks per source location and aggregating the resulting energy, bounce-height, and passage statistics into design-relevant outputs. Both modules use the same underlying rebound formulation for a block striking the slope surface. Velocity is decomposed into components normal and tangential to the local surface at the moment of impact. The post-impact normal component follows a velocity-dependent restitution law. A separate tool, fundamentally different from the trajectory simulators above: a qualitative scoring system for prioritizing which rock slopes most need mitigation, not simulating any individual trajectory. Two selectable methods, both scoring nine site characteristics on an exponential 3/9/27/81-point scale and summing into a total hazard score: RHRS — the original system developed by Pierson et al. (1990) for the Oregon Department of Transportation. All numeric thresholds verified directly against the official FHWA Participant's Manual (Pierson & Van Vickle, 1993, Report FHWA-SA-93-057) rather than secondary sources. mRHRS — Budetta's (2004) modification, confirmed by direct comparison against the original manual to differ in exactly three categories: geologic character (replaced with Romana's Slope Mass Rating in place of the original's two-case structural/erosion assessment), roadway width (different numeric thresholds), and block volume (a per-boulder volume metric, roughly 5× smaller in scale than the original's per-event released-quantity metric). Climate/water and rockfall history use identical criteria in both versions, confirmed against the primary source.

Most rockfall simulators take the block size distribution as a user input. The RBSD module derives it instead from the rock mass itself, following the standard two-stage logic of the rockfall fragmentation literature: an in-situ block size distribution (IBSD) defined by the joint network, transformed into a rockfall block size distribution (RBSD) actually delivered to the slope after kinematic release and fragmentation.

The in-situ block sizes can be obtained two ways, selected by the Block size input mode control:

• DFN-based (the four stages above) — derives the in-situ block size distribution from the joint-set statistics and slope geometry.

• Mapped distribution — the user enters a measured in-situ block-size distribution directly (assumed lognormal, specified by mean and standard deviation of either block volume or diameter, plus a number of blocks or a total detaching volume). The DFN and kinematic stages are skipped and fragmentation applies directly. This suits sites where block size is characterised from scanline/window mapping, photogrammetry, or a volumetric joint count (Jv) rather than an explicit DFN. Because the kinematic release filter cannot run without geometry, the user states whether the entered distribution represents the already-detaching blocks (fragmented directly) or the full in-situ population (a user-supplied release fraction is then applied first). In this mode the blocks have no coordinates, so the release-location coupling options are unavailable.

DFN generation. A discrete fracture network is built from the user's joint sets. Each set is sampled by mean orientation (dip / dip direction) with Fisher-distributed poles, a lognormal disc-radius distribution, and a target volumetric fracture-area density P₃₂ (m²/m³). Fractures are placed until P₃₂ is reached within the slope's bounding domain (Dershowitz & Herda 1992).

Block cutting. The DFN is intersected with the slope face, and removable convex blocks are identified using Goodman & Shi (1985) block theory — a block is kinematically removable if it has finite volume and an empty block pyramid. This yields the in-situ block population and their volumes (the IBSD).

Release filter. Each block is classified by kinematic mode (falling, planar sliding, wedge sliding, or stable) and assigned a factor of safety from limit-equilibrium analysis on its governing joint(s). Blocks are "released" when their mode is removable and FS is below a user threshold. This is the IBSD → released-block step that the simple IBSD overestimates if skipped.

Fragmentation. Released blocks are fragmented on detachment to give the deposited RBSD. Fragmentation follows the Rockfall Fragmentation Distribution approach of Ruiz-Carulla, Corominas & Mavrouli (2017).