DHSVM-MP-TWP

This repository is a research fork of the Distributed Hydrology Soil Vegetation Model (DHSVM) with a microplastics module focused on tire wear particles (TWP). The hydrologic model remains DHSVM; setting PLASTICS = FALSE disables the added particle calculations.

Publications

Li, Z., Huang, Y., Sun, N., Duan, Z., Wigmosta, M., Yang, Z., & Maurer, B. (2026). Traffic-informed prioritization of tire and road wear particle monitoring and mitigation in the Delaware River Basin and Philadelphia metropolitan region. Submitted to Journal of Environmental Management.

TWP additions

The TWP module adds:

• an optional airborne-microplastic deposition pathway;
• a gridded, cell-aggregated annual average daily traffic (AADT) source map;
• antecedent dry-period accumulation and runoff-dependent wash-off;
• division of source mass between impervious and pervious surface stores;
• conservative transfer of mobilized particle mass to the stream network;
• advective channel routing with optional link-to-link diffusion;
• water-column settling, bed deposition, and bed resuspension;
• time-varying TWP source scaling; and
• TWP concentration, load, spatial-state, and mass-balance outputs.

Repository scope

This repository intentionally contains only the core model source. It excludes the test datasets, RBM model, auxiliary conversion programs, legacy GIS scripts, historical tutorials, and generated build products.

Activating the particle module

Set the following in the DHSVM input file:

[OPTIONS]
PLASTICS = TRUE

Add a [PLASTIC] section. TWP MAP FILE activates the TWP pathway. ATMS MAP FILE controls the older airborne-microplastics pathway and can be none.

[PLASTIC]
ATMS MAP FILE = none
TWP MAP FILE = /absolute/path/to/aggregated_aadt.bin
ANTECEDENT DRY DAYS = 7
ATMS MP LOWER THRESH = 0
ATMS MP UPPER THRESH = 0.001
TWP LOWER THRESH = 0.0001
TWP UPPER THRESH = 0.001
TWP SCALE FILE = none
TWP EMISSION FACTOR MG/VKM = 51.1
TWP SOURCE COEFFICIENT = 1
TWP PARTICLE DIAMETER M = 0.000075
TWP SUBMERGED SPECIFIC DENSITY = 0.3
TWP KINEMATIC VISCOSITY M2/S = 0.000001
TWP VON KARMAN CONSTANT = 0.41
TWP REFERENCE HEIGHT RATIO = 0.05
TWP HORIZONTAL DIFFUSIVITY M2/S = 0.1
TWP INITIAL BED MASS KG/M2 = 0

Configuration reference

Key	Unit or format	Default	Purpose
ATMS MAP FILE	DHSVM raster	none	Airborne MP deposition rate in kg m^-2 day^-1.
TWP MAP FILE	DHSVM raster	none	Cell-aggregated AADT in vehicles day^-1. For NetCDF, use variable MP.TWPmp.
ANTECEDENT DRY DAYS	day	0	Initial dry-period source accumulation.
ATMS MP LOWER THRESH	m timestep^-1	0	Runoff depth at or below which airborne MP wash-off is zero.
ATMS MP UPPER THRESH	m timestep^-1	0.001	Runoff depth at or above which airborne MP wash-off is complete.
TWP LOWER THRESH	m timestep^-1	0	Runoff depth at or below which TWP wash-off is zero.
TWP UPPER THRESH	m timestep^-1	0.001	Runoff depth at or above which TWP wash-off is complete. Must exceed the lower threshold.
TWP SCALE FILE	text file or none	none	One nonnegative source multiplier per timestep. The run stops for an invalid or exhausted file.
TWP EMISSION FACTOR MG/VKM	mg vehicle^-1 km^-1	51.1	tire wear emission factor.
TWP SOURCE COEFFICIENT	dimensionless	1	Source adjustment coefficient (f_ct).
TWP PARTICLE DIAMETER M	m	0.000075	Median particle diameter (d50).
TWP SUBMERGED SPECIFIC DENSITY	dimensionless	0.3	(rho_p-rho_w)/rho_w; 0.3 corresponds to 1300 kg m^-3 particles in 1000 kg m^-3 water.
TWP KINEMATIC VISCOSITY M2/S	m^2 s^-1	0.000001	Water kinematic viscosity.
TWP VON KARMAN CONSTANT	dimensionless	0.41	Von Karman constant used in the Rouse number.
TWP REFERENCE HEIGHT RATIO	dimensionless	0.05	Nominal reference height divided by flow depth.
TWP HORIZONTAL DIFFUSIVITY M2/S	m^2 s^-1	0	Stream-link longitudinal diffusivity; zero disables diffusion.
TWP INITIAL BED MASS KG/M2	kg m^-2	0	Initial re-entrainable TWP bed mass for every stream segment.

Preparing the TWP map

The TWP raster represents cell-aggregated AADT, not a point-station table:

1. Project traffic-count locations into the DHSVM grid coordinate system.
2. Assign every station to a DHSVM cell.
3. Sum AADT where multiple stations fall in the same cell.
4. Write a raster with the dimensions, orientation, and format of the other DHSVM input maps.
5. Use zero for cells without a source; do not use missing values inside the modeled basin.
6. For temporal variation, provide one nonnegative scale value for every model timestep.

The source conversion is:

P_TWP = f_ct E AADT DX / 1000
q_TWP = P_TWP 1e-6 / 86400

Here P_TWP is mg day^-1, q_TWP is kg s^-1, E is the emission factor in mg vehicle^-1 km^-1, and DX is grid spacing in meters. The temporal scale factor is applied to q_TWP each timestep.

Wash-off is linear between the configured runoff thresholds:

F = 0                                      IExcess <= lower
F = (IExcess-lower)/(upper-lower)          lower < IExcess < upper
F = 1                                      IExcess >= upper

Channel transport

Hydrologic network routing supplies advection. Optional link-to-link diffusion uses:

M_diff = epsilon_s A (C_up-C_down) Dt / L

The stream segment is treated as rectangular. Flow depth is the greater of storage depth and the Manning normal depth calculated from routed flow, width, slope, and roughness. Particle settling velocity is evaluated from particle diameter, submerged density, and water viscosity. The water-column/near-bed concentration ratio uses a numerically integrated Rouse profile.

The equilibrium reference concentration follows the van Rijn suspended-load formulation. Water-column/bed exchange is:

Delta M = ws (Ceq-Ca) width length Dt

Positive exchange entrains stored bed mass; negative exchange deposits suspended mass. Entrainment is limited by available bed mass and deposition by available suspended mass, preserving mass. The formulation follows L. C. van Rijn (1984), Journal of Hydraulic Engineering 110(11), 1613-1641, DOI 10.1061/(ASCE)0733-9429(1984)110:11(1613).

TWP outputs

File or variable	Quantity	Unit
TWP.Load.Only	TWP mass leaving each recorded stream segment per timestep	kg
TWP.Conc.Only	Mixed depth-averaged TWP concentration	kg m^-3
MP.TWPmp	Input TWP/AADT raster variable	vehicles day^-1 per cell
MP.mp_accum	Accumulated surface microplastic state	model state unit

For concentration conversion, 1 kg m^-3 = 1000 mg L^-1.

The DHSVM mass-balance output also includes particle input, surface storage, channel export, and residual/error terms. Review the residual when validating a new setup.

Constraints and current limitations

• TWP UPPER THRESH must be greater than TWP LOWER THRESH.
• Particle diameter, submerged density, viscosity, and the von Karman constant must be positive.
• The reference-height ratio must be between zero and one.
• Diffusivity and initial bed mass cannot be negative.
• TWP map values and temporal scale factors cannot be negative.
• A complete TWP simulation requires a routed stream network. Unit-hydrograph mode does not perform channel particle routing.
• The TWP implementation is a research extension and should be checked against field data and independent mass-balance calculations before operational use.

License and attribution

Retain the upstream DHSVM attribution and license terms that apply to this source. Publications using this fork should cite DHSVM, the relevant TWP or microplastics methodology, and the exact Git commit or release tag used.