Note This is a placeholder for CCMMF milestones; the spec has been migrated to the SIPNET repository PecanProject/sipnet#75

Summary

The current implementation of moisture dependence functions in SIPNET needs to be revised and expanded to properly represent:

1. RH Moisture Dependence decline at high soil moisture (under anaerobic conditions). @jmzobitz suggested using an empirical equation he used in Zobitz et al. 2021 that comes from Moyano et al. 2013
2. N₂O Moisture Dependence Currently proposed as beta function Properly represent N₂O fluxes with a unimodal response to soil moisture.
3. Add an empirical function for CH₄ flux.

Proposed Changes

1. Replace the linear function (Eq. 24) with the Moyano et al. (2013) empirical function

- **Current Function**
  $$
  D_{\text{water}_{R_H}} =
  \begin{cases}
      1, & \text{if } T_{\text{soil}} \leq 0 \\
      f_{\text{WHC}}, & \text{if } T_{\text{soil}} > 0
  \end{cases}
  $$
- **Proposed function**:
  $$
  D_{\text{water}_{R_H}} =
  \begin{cases}
      1, & \text{if } T_{\text{soil}} \leq 0 \\
      3.11 f_{\text{WHC}} - 2.42 f_{\text{WHC}}^2, & \text{if } T_{\text{soil}} > 0
  \end{cases}
  $$

library(ggplot2)

# Wang et al 2023
f_aerobic <- function(f_WHC) 3.11 * f_WHC - 2.42 * f_WHC^2

# Truncated Beta PDF anaerobic metabolism
dbeta_scaled <- function(f_WHC, a = 10, b = 1.5, f_WHC_min = 0.5, f_WHC_max = 1) {
  db <- dbeta((0:1000/1000), a, b)
  zeros <- (0:1000/1000) < f_WHC_min
  dbs <- ifelse(zeros, NA, db / max(db))
  return(dbs)
}

# Create data frames for ea. fn
df <- data.frame(f_WHC = seq(0, 1, length.out = 1001))
df$D_anaerobic <- dbeta_scaled(df$f_WHC)
df_linear <- data.frame(f_WHC = df$f_WHC, D_water = df$f_WHC)
df$D_aerobic_and_anaerobic <- pmin(df_linear$D_water, df$D_anaerobic, na.rm = TRUE)


cb_palette <- c(
  "Linear Scaling: f_WHC" = "#E69F00",
  "Wang et al 2023\n3.11 * f_WHC - 2.42 * f_WHC^2" = "#56B4E9"#,
  #"Anaerobic (current): \nTruncated Beta(10,1.5) uncalibrated example" = "#009E73",
  #"Minimum of Aerobic & Anaerobic" = "black"
)

ggplot() + 
  geom_line(data = df_linear, aes(x = f_WHC, y = D_water, 
                                  color = "Linear Scaling: f_WHC")) +
  geom_line(data = df, aes(x = f_WHC, y = f_aerobic(f_WHC), 
                           color = "Wang et al 2023\n3.11 * f_WHC - 2.42 * f_WHC^2")) +
  
  # Anaerobic function (commented out from legend and plot)
  # geom_line(data = df, aes(x = f_WHC, y = D_anaerobic),
  #           color = "#009E73") +
  
  # Combined aerobic and anaerobic function as a black dotted line
  # geom_line(data = df, aes(x = f_WHC, y = D_aerobic_and_anaerobic, 
  #                          color = "Minimum of Aerobic & Anaerobic"),
  #            linetype = "dotted") +
  
  scale_color_manual(name = "function", values = cb_palette) +
  
  # Set axis labels and limits
  scale_y_continuous(limits = c(0, 1), breaks = seq(0, 1, 0.1)) +
  xlab("Fraction of WHC") + 
  ylab("D_moisture") +
  ggtitle("RH Moisture Dependence Functions") +
  
  theme_minimal() +
  theme(legend.position = c(0.75, 0.2),
        legend.background = element_rect(fill = "white"))

2. Update GHG flux moisture dependence functions; perhaps separate empirically fit functions for CH4 and N2O production

• For N2O, Wang et al 2023 propose a Gaussian distribution over the range of soil moisture [0,1.4]
• CH4 Modeled in DayCent as a beta distribution.

Challenges

The current RH moisture dependence function is linear - equal to soil moisture as a fraction of water holding capacity. The proposed one is quadratic. This representation has long worked for (almost exclusively) temperate forests where soils are rarely and if so, only temperarily, saturated.

Changing this functional form could affect other users, so should be added with an optional flag.

References

• Del Grosso, S. J. D., et al. (2000), General CH4 oxidation model and comparisons of CH4 Oxidation in natural and managed systems, Global Biogeochem. Cycles, 14(4), 999–1019, doi:10.1029/1999GB001226.
• Moyano, F. E., Manzoni, S., & Chenu, C. (2013). Responses of soil heterotrophic respiration to moisture availability: An exploration of processes and models. Soil Biology and Biochemistry, 59, 72–85.
• Zobitz, J. et al. (2021). Comparing an exponential respiration model to alternative models for soil respiration components in a Canadian wildfire chronosequence. Geosci. Model Dev., 14, 6605–6622. DOI
• Wang H. et al. (2023). Quantifying nitrous oxide production rates from nitrification and denitrification under various moisture conditions in agricultural soils. Front. Microbiol., 13:1110151. DOI