Lake Water
Temperature Tracking

This uncalibrated pilot estimates conductive bottom growth using a Stefan/Ashton-style resistance model: existing ice, accumulating snow and surface-to-air heat exchange. Hourly NWS temperature and wind drive the calculation. NWS snowfall totals are distributed evenly within each forecast interval. Both starting snow-depth scenarios receive the same accumulation; all new snow is assumed retained. The two scenarios use maximum snow/minimum wind and minimum snow/maximum wind. They are not statistical uncertainty bounds.

h_next = √[(h + b)² + 2 kᵢ max(0, −T_air) Δt / (ρᵢ L)] − b
b = kᵢ (starting_snow / k_existing + added_snow / k_fresh + 1 / H_air)

Ice conductivity 2.24 W/(m·K), density 917 kg/m³, latent heat 334,000 J/kg. Snow conductivity is 3.2 × 10⁻⁶ × density². Existing snow uses your selected density; new snowfall is a separate light-snow layer at an assumed 180 kg/m³. Mid-hour accumulated snow is used for each growth step. This is an assumption, not a measured snow density. As a pilot assumption, surface heat transfer increases linearly from 10 W/(m²·K) at calm wind to 25 at 4 m/s and is capped there. This wind interpolation is not a validated local calibration.

Above-freezing hours contribute zero growth; melt is not modeled or subtracted. This is gross freezing potential, not net change or final ice thickness. Sunlight, water heat, currents, bubblers, radiation, rain, slush, flooding, snow settling, drifting, snowmelt, wind damage and ice strength are not resolved. Local forecasts may differ from the ice surface. The calculation stops at the end of overlapping temperature, wind and snowfall coverage. Missing snowfall is never treated as zero. Results can overstate actual growth. Thickness inputs are unverified estimates at the calculation start time. Do not use these numbers to decide whether to travel on ice.

References: Stefan growth theory; Comfort & Abdelnour: snow insulation and surface heat transfer. NH ice-safety guidance.