CLI reference: lsdtt-cosmo-tool

Inverts measured cosmogenic nuclide concentrations to basin-averaged effective erosion rates using the production–accumulation–inversion approach. Loads a DEM, builds D8 flow routing, computes a Stone 2000 production raster, snaps sample points to the channel network, and writes a CSV of erosion rate estimates with optional uncertainty bounds.

The DEM must be in a projected CRS (e.g. UTM). At least one sample must be successfully snapped to a channel.

Usage

lsdtt-cosmo-tool <path> <parameter_file>
lsdtt-cosmo-tool <parameter_file>         # path defaults to ./
lsdtt-cosmo-tool -h                       # print help and exit
lsdtt-cosmo-tool -v                       # print version and exit

Parameter file format

# Lines beginning with # are comments.
key: value

Pipeline overview

Step	Description
1	Load DEM; check projected CRS
2	Pit fill (skipped when `fill_before_routing: false`)
3	D8 flow routing
4	Pressure raster (from `atmosphere`), then the production raster — Stone 2000 or LSDn per `scaling_scheme`
5	Load sample points from CSV; snap to channel
6	For each sample: apply standardisation, invert to erosion rate, compute uncertainty bounds
7	Write output CSV

Input / output parameters

Parameter	Type	Default	Description
`read_path`	string	directory of parameter file	Directory containing the input DEM.
`read_fname`	string	parameter file stem	Input raster filename (without extension).
`write_path`	string	same as `read_path`	Directory for output files.
`write_fname`	string	same as `read_fname`	Output filename stem.

The output CSV is always written as <write_fname>_CN_ERATE.csv.

CRN physics

Parameter	Type	Default	Description
`nuclide`	string	`Be10`	Nuclide used for erosion rate inversion. Options: `Be10`, `Al26`, `C14`.
`muon_scheme`	string	`BraucherBorchers`	Muon production parameter scheme. Options: `BraucherBorchers`, `Braucher`, `Granger`, `Schaller`, `newCRONUS`, `cronusEmulatorSt`, `cronusEmulatorLSDn`. The two `cronusEmulator` schemes fit lsdtt3's three muon pathways to the CRONUS v3 tabulated* model-1A muon inventory — use them when the goal is agreement with the CRONUS v3 online calculator (measured mean −0.23 % on Be-10; see the benchmarking doc). They are pressure-dependent, so a pressure raster is computed automatically, and nuclide-specific. Pair `cronusEmulatorSt` with `scaling_scheme: stone` and `cronusEmulatorLSDn` with `scaling_scheme: lsdn`; the pairing is enforced by the spallation reference production, which differs between the two. `cronusEmulatorLSDn` carries a larger residual than the St variant for structural reasons documented in the benchmarking doc.
`rock_density_kg_m3`	float	`2650`	Rock bulk density (kg/m³). Used to convert effective erosion rate (g/cm²/yr) to mm/kyr.
`scaling_scheme`	string	`stone`	Production-rate scaling. `stone` = Stone (2000)/Lal, the default and the CAIRN/OCTOPUS scheme. `lsdn` = Lifton, Sato & Dunai (2014) nuclide-specific scaling, interpolated from the LSDn2015 grids; needs the LSDn grids in `crn_parameters.nc` and supports `Be10`, `Al26`, `C14` only. Both are evaluated per pixel, including when `sample_age_yr > 0`. `muon_scheme` is unaffected.

Atmosphere and palaeomagnetic correction

Parameter	Type	Default	Description
`crn_nc_fname`	string	``	Path to `crn_parameters.nc` (ERA40 grids + Muscheler VDM record), for the ERA40 atmosphere and palaeomagnetic correction. Leave empty to use the `crn_parameters.nc` bundled with lsdtt3 (found automatically relative to the executable); if that is missing too, the standard (SLHL) atmosphere is used with no correction.
`sample_age_yr`	float	`0`	Sample exposure age (yr BP). Drives the time-integrated production correction; what it selects depends on `scaling_scheme`. Under `stone` it is the Muscheler VDM palaeomagnetic correction; under `lsdn` it is the duration-weighted average of LSDn production over the Pavón-Carrasco record, so it is longitude-dependent as well as latitude-dependent. `0` = modern field, no correction — the default, so no time-series correction is applied unless you ask for it; see below. Requires the record in `crn_parameters.nc`. "Now" is taken as 2010, the reference year of the packed record.
`atmosphere`	string	`ERA40`	`ERA40` (Balco/Lifton climatology + latitude-dependent lapse, the lsdtt3 default), `NCEP` (the atmosphere CAIRN/OCTOPUS use; needs the NCEP grids), or `standard` (SLHL barometric formula). `ERA40`/`NCEP` fall back to `standard` if their grids are absent.
`self_consistent_age`	bool	`false`	Solve the sample age and the erosion rate together, iterating to the erosional integration timescale `t = Γ_sp/ε` rather than holding the age at `sample_age_yr`. This is the CRONUScalc / Balco v3 convention, and it is how most published LSDn rates were computed — see below. Leaving it `false` while `muon_scheme: cronusEmulatorLSDn` raises an advisory warning, because that combination cannot reproduce CRONUS v3: measured bias at Sulawasi is +10.9 % at `sample_age_yr: 0` against +0.5 % with the iteration on. The warning is advisory — a fixed-age LSDn run is perfectly legitimate, it is just not the route to v3 agreement.
`self_consistent_age_tolerance`	float	`0.001`	Relative convergence tolerance on ε for the iteration. The default sits just inside the inversion solver's own round-trip tolerance, so tightening it buys precision the forward model cannot deliver.
`self_consistent_age_max_iter`	int	`10`	Iteration cap. A sample that has not converged is reported with a warning and flagged `self_consistent_converged = 0` in the output CSV; its last iterate is written, never silently accepted.

No time-series correction is applied by default

Time-integrated production is opt-in, not the default. It is switched on by either sample_age_yr > 0 or self_consistent_age: true, and both are off out of the box. With the defaults, production is evaluated at the present-day field under either scaling scheme — stone applies no Muscheler VDM correction, and lsdn reads record step 0 with no time loop. Every run reports which it did:

Palaeomag     : off (modern field)

Check that line if you are unsure what a given parameter file produced.

CAIRN-comparable run

CAIRN applied no palaeomagnetic time-series correction, so reproducing that convention means leaving the time settings alone and matching the scaling and atmosphere:

muon_scheme:         Braucher   # CAIRN's default muon scheme
scaling_scheme:      stone      # Stone (2000)/Lal — the CAIRN/OCTOPUS scaling
atmosphere:          NCEP       # the atmosphere CAIRN/OCTOPUS use
sample_age_yr:       0          # default — modern field, no correction
self_consistent_age: false      # default — no iteration to t = Γ_sp/ε
topo_shield_value:   1          # default — see the shielding caveat below

Two of these differ from the lsdtt3 defaults: muon_scheme (lsdtt3 defaults to BraucherBorchers) and atmosphere (lsdtt3 defaults to ERA40). Both change the answer on their own, so set both explicitly for a like-for-like comparison. The atmosphere difference is a pressure difference, not a time-series one.

Topographic shielding is the one convention lsdtt3 deliberately does not follow. CAIRN applies a topographic shielding correction; lsdtt3 defaults to topo_shield_value: 1, i.e. none, following DiBiase (2018) — see Why topographic shielding defaults to off. This is a deliberate methodological difference from CAIRN, not an omission, and it is why a CAIRN-convention run and an lsdtt3 default run will not agree exactly even with everything above matched. Set topo_shield_value to a basin shielding factor if you want CAIRN's convention.

Finally, this note fixes the convention, not the numerical agreement: for validated per-sample targets against the CAIRN datasets (Bierman 2005, Dethier 2014, Kirchner 2001, Munack, …) see Benchmarking and testing in lsdtt3, which also records that CAIRN used SRTM 90 m throughout, so DEM resolution is a further difference to control.

Self-consistent age

It applies to both scaling schemes, not just LSDn. The iteration rebuilds the production raster at the converged age whatever scaling_scheme is set to, and what the age changes is scheme-specific — under lsdn it is the duration-weighted LSDn average, under stone it is the Muscheler VDM palaeomagnetic correction. Measured on the bundled tiny fixture, turning it on moves the rate +9.0 % under stone (19.96 → 21.76 mm/kyr) and +8.7 % under lsdn (17.72 → 19.26 mm/kyr).

The one case where it does nothing is when crn_parameters.nc is unavailable: the palaeomagnetic correction needs the VDM record, so with no NetCDF file the age has no effect under stone and the iteration converges immediately at the modern-field rate. (lsdn fails outright without the file, rather than falling back.)

A sample's nuclide inventory accumulates over roughly one attenuation length of erosion, τ = Γ_sp/ε (Γ_sp = 160 g/cm² in every muon scheme lsdtt3 ships). The geomagnetic field it actually integrated is therefore the field over that window — but τ depends on ε, the very thing being solved for. With self_consistent_age: true the tool closes the loop: invert at sample_age_yr, form t = Γ_sp/ε, rebuild production, re-invert, repeat until ε stops moving. sample_age_yr becomes only the starting guess (0 = start from the modern field). Typically 2–4 iterations.

It is off by default for two reasons. It changes the meaning of the reported rate, so it should be a deliberate choice; and it costs one production-raster rebuild per iteration per sample — the age, and hence the whole raster, differs sample to sample. With scaling_scheme: lsdn on a large DEM this dominates the run time.

Turn it on when comparing against published LSDn or CRONUScalc rates, which are computed this way — but note that "computed this way" is a property of the reference calculator, not of the scaling scheme, so the same argument applies to a stone run being compared against CRONUScalc. The effect is largest where the past field departed furthest from modern — low latitudes, and slowly eroding samples with long τ.

Three extra columns appear in the output CSV: self_consistent_t_yr (the converged age), self_consistent_n_iter, and self_consistent_converged (1/0). Uncertainty bounds are inverted against the central sample's converged production raster, matching how the reference calculators propagate concentration uncertainty with scaling held fixed.

Sample points

Parameter	Type	Default	Description
`sample_points_fname`	string	`CRN_points.csv`	CSV file containing sample locations and concentrations. Must have `latitude` and `longitude` columns (EPSG:4326).
`sample_column_name`	string	`sample_name`	Column name for sample identifiers. Optional; omitted from output if not present.
`concentration_column_name`	string	``	Column name for measured nuclide concentration (atoms/g). Defaults to `<nuclide>_CONC` if not set.
`uncertainty_column_name`	string	``	Column name for 1-sigma concentration uncertainty (atoms/g). Defaults to `<nuclide>_uncertainty`. If absent, uncertainty bounds are skipped.
`standardisation_column_name`	string	`standardisation`	Column name for AMS standardisation name. If absent, no standardisation correction is applied. Supply the concentration as measured on its own standard — see the warning below.

⚠ Standardisation is applied by lsdtt3 — do not pre-normalise

When standardisation_column_name names a column that is present and non-empty, lsdtt3 multiplies both the concentration and its uncertainty by the AMS standardisation factor for that standard (e.g. Be-10 KNSTD → 07KNSTD, factor 0.9042). The concentration_* value written to the output CSV is the normalised one.

Two consequences, and both have caused real errors:

1. Give the tool concentrations as your lab reported them. If you normalise to 07KNSTD yourself and leave the standardisation column populated, the factor is applied twice. A 10 % error in concentration propagates straight into the erosion rate. To supply already-normalised values, either drop the column or set standardisation_column_name to a name that does not exist in the file.

2. Do not feed the output CSV's concentration_* back into another calculator together with the standard name. It is already normalised; pairing it with the standard name asks the second calculator to normalise it again. This is exactly what happened in the OCTOPUS comparison tier — the concentrations arrived already on the 07KNSTD basis and were corrected a second time, which inverted the apparent sign of the residual (clapp2002 read +6.6 % instead of −4.09 %) and sent the investigation after a non-existent physics discrepancy. See Benchmarking and testing in lsdtt3.

The CRONUS v3 input file that this tool writes handles the second case correctly: it carries the as-measured concentration with the standard name, so the calculator applies the factor exactly once.

Known exception: tests/scripts/compare_cosmo_tool_vs_cronus_v3.py reads the normalised concentration_* from the output CSV, so passing it --standardisation-col reintroduces the double-count. Prefer the CLI's own <write_fname>_CRONUS_V3_INPUT.txt.

Uncertainty propagation

The internal/external split follows the CRONUS v3 convention. CRONUS reports an internal (analytical, measurement-only) uncertainty and an external uncertainty that additionally carries the production-rate calibration and decay-constant terms. lsdtt3 uses the same two-tier structure and the same names, so apparent_e_unc_int_mm_kyr is comparable to the CRONUS internal uncertainty and apparent_e_unc_ext_mm_kyr to the external one.

The consequence is that lsdtt3's unc_ext is only comparable to a CRONUS external uncertainty once you switch the extra terms on — see the two propagate_* parameters below. Left at their defaults, unc_ext equals unc_int and both are concentration-only, so quoting unc_ext as an external uncertainty would understate it. This is the single most common way to under-report absolute uncertainty from this tool.

Uncertainty bounds are only produced when the input CSV carries the column named by uncertainty_column_name (default <nuclide>_uncertainty). The concentration sigma is scaled by the same AMS standardisation factor as the concentration itself before it is used — so it must be supplied as measured, on the same basis as the concentration. See the standardisation warning under Sample points.

The internal uncertainty is a re-inversion at each concentration bound: unc_int = ½ |ε(C+σ) − ε(C−σ)|. Both bounds are inverted against the central sample's production raster, with scaling held fixed — this matches how the reference calculators propagate concentration uncertainty, and keeps the bounds an uncertainty on the concentration rather than a second coupled solve.

The external uncertainty adds the terms below in quadrature: unc_ext = √(unc_int² + (ε·production_rel_sigma)² + (ε·decay_rel_sigma)²).

Both switches are off by default, so unc_ext equals unc_int unless you turn them on. Reported uncertainties are then concentration-only and are an underestimate of the absolute rate uncertainty — enable these for publishable absolute uncertainties.

Parameter	Type	Default	Description
`propagate_production_uncertainty`	bool	`false`	Add a production-calibration term to the external erosion-rate uncertainty, using `production_rel_sigma`.
`production_rel_sigma`	float	`0`	Relative 1-sigma production uncertainty (dimensionless, e.g. `0.05` for 5 %). Used only when `propagate_production_uncertainty: true`. Must be ≥ 0 or the run exits fatally.
`propagate_decay_uncertainty`	bool	`false`	Add a decay-constant term to the external erosion-rate uncertainty, using `decay_rel_sigma`.
`decay_rel_sigma`	float	`0`	Relative 1-sigma decay uncertainty term (dimensionless). Used only when `propagate_decay_uncertainty: true`. Must be ≥ 0 or the run exits fatally.

Shielding

Shielding values are applied uniformly across the entire basin.

Parameter	Type	Default	Description
`self_shield_eff_thickness`	float	`0`	Sample self-shielding as effective depth (g/cm²). `0` = surface sample.
`snow_shield_eff_thickness`	float	`0`	Snow shielding as effective depth (g/cm²). `0` = no snow correction.
`topo_shield_value`	float	`1`	Topographic shielding factor (dimensionless, 0–1). `1` = no topographic shielding — the default, deliberately; see below.

Why topographic shielding defaults to off

lsdtt3 applies no topographic shielding correction by default, which is a deliberate departure from CAIRN and from most older basin-averaged workflows.

The reasoning follows DiBiase (2018). The conventional correction accounts only for the reduction of production at the surface, while assuming the effective mass attenuation length is spatially uniform. It is not: attenuation length increases on steep slopes, and the two effects largely offset. Correcting for one and not the other introduces a slope-dependent systematic error — which is worse than a bias, because slope correlates with erosion rate, so it can manufacture spurious topography-erosion relationships. In steep landscapes the surface-only correction reaches ~20 %.

DiBiase's conclusion is that topographic shielding corrections are inappropriate for catchment-mean erosion rates in most cases, and are needed only for steep catchments with a non-uniform distribution of quartz and/or erosion rate. That is the case in which to set topo_shield_value yourself — along with reproducing a published rate computed the older way, such as a CAIRN rate.

Two things to know if you do set it: it is applied as a single uniform value across the whole basin, not per pixel; and in lsdtt3's published-dataset comparison tier the topographic shielding term is decomposed at roughly +0.70 % at basin scale (Benchmarking and testing in lsdtt3).

DiBiase, R. A. (2018). Short communication: Increasing vertical attenuation length of cosmogenic nuclide production on steep slopes negates topographic shielding corrections for catchment erosion rates. Earth Surface Dynamics, 6, 923–931. https://esurf.copernicus.org/articles/6/923/2018/

The same argument does not apply to self_shield_eff_thickness or snow_shield_eff_thickness, which are ordinary depth-attenuation terms and default to 0 simply because the common case is an unshielded surface sample.

DEM preparation

Parameter	Type	Default	Description
`fill_before_routing`	bool	`true`	Fill DEM pits before building D8 flow routing. When `false`, route the input DEM unchanged; use this for an already filled DEM or when sinks must intentionally be preserved.
`min_slope_for_fill`	float	`0.0001`	Minimum imposed slope during pit filling (m/m). Used only when `fill_before_routing: true`.
`boundary_conditions`	string	`oooo`	D8 flow routing boundary conditions. Four-character string (N,E,S,W): `o` = open, `b` = forced outlet.

Channel snapping

Sample points are snapped to the nearest channel pixel meeting the contributing-pixel threshold using an expanding square search.

Parameter	Type	Default	Description
`threshold_contributing_pixels`	int	`1000`	Minimum contributing pixels for a node to be considered a channel.
`sample_snap_max_search_radius_pixels`	int	`8`	Maximum search radius (pixels) when snapping points to a channel.

Output

Parameter	Type	Default	Output file
`write_production_raster`	bool	`false`	`<write_fname>_PROD.tif` — production scaling raster (Stone 2000 or LSDn, per `scaling_scheme`).
`write_basin_production_csv`	bool	`false`	`<write_fname>_BASIN_PROD.csv` — per-sample basin-mean production scaling and mean elevation.
`write_cronus_v3_input`	bool	`true`	`<write_fname>_CRONUS_V3_INPUT.txt` — a ready-to-paste CRONUS v3 batch request. See below.

The erosion rate CSV <write_fname>_CN_ERATE.csv is always written. Columns include:

sample metadata (sample_name when present, nuclide),
snapped outlet location (latitude, longitude),
basin comparison metadata (effective_pressure_hpa, centroid_latitude_deg),
measured + predicted concentration,
effective erosion rate (apparent_eff_e_g_cm2_yr),
erosion rate in mm/kyr (*_mm_kyr).

For backward compatibility, duplicated legacy *_mm_yr aliases are still written and carry the same mm/kyr numeric values.

When an uncertainty column is present in the input, these are appended: the lower and upper bounds (concentration_minus_*, predicted_conc_minus_*, apparent_eff_e_minus_g_cm2_yr, apparent_e_minus_mm_kyr and the matching *_plus_* set), and the summary uncertainties apparent_eff_e_unc_int_g_cm2_yr / apparent_e_unc_int_mm_kyr and apparent_eff_e_unc_ext_g_cm2_yr / apparent_e_unc_ext_mm_kyr.

With self_consistent_age: true, three further columns appear: self_consistent_t_yr, self_consistent_n_iter, self_consistent_converged.

Basin production diagnostic

write_basin_production_csv is a reporting-only diagnostic — it has no effect on the inversion. For each snapped sample it averages the production scaling raster and the DEM elevation over the exact upslope pixel set the inversion uses, so the basin-mean production can be checked against an external oracle such as OCTOPUS BEPROD/ELEV_AVE. Columns: sample_name, snapped_latitude, snapped_longitude, n_upslope_pixels, mean_prod_scaling, mean_elev_m.

lsdviztools3 interface notes

lsdviztools3 can plot comparison figures from the standard cosmogenic CSV workflows that include lsdtt-cosmo-tool outputs.

Basin/sample comparison tiers (run_cosmo_validation.py, run_cosmo_riversand_comparison.py) produce summary CSVs under tests/expected/cosmo/published/ such as:
<dataset>_summary.csv
<dataset>_lsdn_summary.csv
<dataset>_riversand_summary.csv
lsdviztools3 command:

lsdviz3-cosmo-published \
  --published-dir tests/expected/cosmo/published \
  --datasets dethier2014,clapp2002,dibiase2010 \
  --mode auto \
  --outdir demo_output/cosmo

For sweep-style lsdtt3-vs-reference figures, see the single-point scan script tests/scripts/run_cosmo_single_point_scaling_scan.py and lsdviz3-cosmo-compare.

For sample-set comparison against CRONUS v3 (separate from synthetic point scan), use:

python3 tests/scripts/compare_cosmo_tool_vs_cronus_v3.py \
  --cosmo-csv <write_path>/<write_fname>_CN_ERATE.csv \
  --points-csv <input_points.csv> \
  --uncertainty-col <sigma_column> \
  --standardisation-col <standardisation_column> \
  --scheme St \
  --out-csv <write_path>/<write_fname>_CRONUS_COMPARE.csv

This builds a CRONUS v3 batch request from each sample's nuclide, concentration, basin-effective pressure, and basin-centroid latitude, then writes per-sample lsdtt3-vs-CRONUS erosion-rate residuals.

Building a CRONUS v3-compatible input file

The CLI writes this file for you — no Python required. write_cronus_v3_input defaults to true, so every run produces <write_fname>_CRONUS_V3_INPUT.txt alongside the erosion-rate CSV. Set it to false to suppress it.

A basin-averaged sample has no single site elevation, so the correct way to present it to a point calculator is as the basin-effective pressure together with the basin centroid latitude — exactly what CRONUS v3's pre (pre-calculated pressure) input mode takes. The tool has both numbers already; they are also reported in the output CSV as effective_pressure_hpa and centroid_latitude_deg.

The file contains two lines per sample:

S00001 0.000027 -7.488704 1002.3418 pre 0 2.65 1 0.0001 2010 ;
S00001 Be-10 quartz 100000.000000 3000.000000 07KNSTD ;

The first line is the sample/site record, the second the nuclide record:

line	fields
site	`id`, latitude, longitude, pressure (hPa), `pre` (flag: the preceding value is a pressure, not an elevation), sample thickness (`0`), density (`2.65`), shielding (`1`), erosion-rate guess (`0.0001`), sample year (`2010`)
nuclide	`id`, nuclide (`Be-10` / `Al-26` / `C-14`), target mineral (`quartz`), concentration (atoms/g), 1-sigma uncertainty (atoms/g), AMS standard (`07KNSTD`, `KNSTD`, …)

Details worth knowing:

Shielding is written from topo_shield_value, so it is 1 (none) by default, matching the DiBiase (2018) reasoning above. Change topo_shield_value and the file follows; the run prints a note when it is not 1.
Density is written from rock_density_kg_m3, converted to g/cm³.
Concentrations are written as measured, not standardisation-normalised, and are paired with the sample's standardisation name so the calculator applies its own factor. lsdtt3 normalises internally, so writing the normalised value and the standard name would apply the correction twice — the error that once produced a spurious sign flip in the OCTOPUS comparison tier.
Sample names are sanitised, since CRONUS ids must be whitespace-free and ; terminates a record: anything outside [A-Za-z0-9_-] becomes _. Duplicate names gain a _2, _3, … suffix, because the calculator keys its results by id and would otherwise merge them.
If the input has no uncertainty column, sigma is written as 0 and the run says so. It is never replaced by a guessed percentage — a fabricated uncertainty in a file you might publish is worse than an obvious zero.
Samples whose pressure or centroid latitude did not resolve are omitted with a warning rather than written as nan, which the calculator would reject.

Sending it to the online calculator

Paste the file's contents into the text box on the CRONUS v3 erosion-rate input page (hess.ess.washington.edu, with a mirror at stoneage.hzdr.de) and submit. The block is already in the calculator's native batch format, so no reformatting is needed.

The online calculator applies its own scaling and muon treatment, so a matching input file is necessary but not sufficient for agreement — see the CRONUS v3 example parameter file below for the lsdtt3 settings that target it.

Automating the comparison

tests/scripts/compare_cosmo_tool_vs_cronus_v3.py submits a batch and writes per-sample residuals, for when you want the round trip scripted rather than manual:

python3 tests/scripts/compare_cosmo_tool_vs_cronus_v3.py \
  --cosmo-csv   <write_path>/<write_fname>_CN_ERATE.csv \
  --points-csv  <input_points.csv> \
  --uncertainty-col     <sigma_column> \
  --standardisation-col <standardisation_column> \
  --scheme St \
  --out-csv <write_path>/<write_fname>_CRONUS_COMPARE.csv

Match --scheme to scaling_scheme (St for stone, LSDn for lsdn). Add --no-submit --text-block-out <file> to build a request without contacting the calculator, though for that the CLI's own output file is usually what you want.

Example parameter file

read_path:  /data/dems/
read_fname: my_dem
write_path: /data/output/
write_fname: my_samples

nuclide:       Be10
muon_scheme:   BraucherBorchers
rock_density_kg_m3: 2650

fill_before_routing:             true
min_slope_for_fill:              0.0001
boundary_conditions:             oooo
threshold_contributing_pixels:   1000
sample_snap_max_search_radius_pixels: 8

sample_points_fname:             CRN_points.csv
concentration_column_name:       Be10_CONC
uncertainty_column_name:         Be10_uncertainty
standardisation_column_name:     standardisation
sample_column_name:              sample_name

write_production_raster: false

Example parameter file: agreement with the CRONUS v3 online calculator

Use this pairing when the goal is to reproduce CRONUS v3. Keep sample_age_yr: 0, pair cronusEmulatorSt with scaling_scheme: stone, and match --scheme St on tests/scripts/compare_cosmo_tool_vs_cronus_v3.py.

read_path:  /data/dems/
read_fname: my_dem
write_path: /data/output/
write_fname: my_samples

nuclide:        Be10
muon_scheme:    cronusEmulatorSt
scaling_scheme: stone
atmosphere:     ERA40
sample_age_yr:  0

sample_points_fname:         CRN_points.csv
uncertainty_column_name:     Be10_uncertainty
standardisation_column_name: standardisation

propagate_production_uncertainty: true
production_rel_sigma:             0.05

For the LSDn route use muon_scheme: cronusEmulatorLSDn with scaling_scheme: lsdn and self_consistent_age: true — see the self_consistent_age row above for why the fixed-age variant will not reproduce v3.

Parameters

This section is generated from the executable's -h json registry. Do not edit it manually.

Status: stable.

File I/O

Parameter	Type	Default	Role	Parent	Description	Guidance
`read_path`	`string`	`./`	`io`	-	Directory from which input files are read.	Defaults to the directory holding the parameter file.
`read_fname`	`string`	-	`io`	-	Input filename, with or without extension (e.g. my_dem.tif).	Read from read_path. lsdtt3 uses GDAL, so any GDAL-readable raster format works. If empty, defaults to the parameter filename stem.
`write_path`	`string`	-	`io`	-	Directory to which output files are written.	Defaults to read_path when not set.
`write_fname`	`string`	-	`io`	-	Output filename prefix (no path, no extension).	All outputs are named <write_fname>_<descriptor>.<ext>. Defaults to the stem of read_fname when not set.