BASEmetab_script/ebase_model.txt at master · fawda123/BASEmetab_script · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
## JAGS model

	model{

    # -------
    # PRIORS

    # set priors
    # amean <- 0.2 / nstepd
    # bmean <- 0.251 / (100 * 3600 / interval)

    # a ~ dunif(arng[1] * amean, 2 * arng[2])
    # r ~ dnorm(0, 1 / rvar)T(0,)
    # b ~ dunif(brng[1] * bmean, brng[2] * bmean) # note that Wanninkhof reports uncertainty of this parameter at +/- 20%

    # set priors
    amean <- 0.2 / nstepd
    asd <- 0.01 * amean
    atau <- 1 / (asd * asd)
    rmean <- 20 / nstepd
    rsd <- 1.341 * rmean
    rtau <- 1 / (rsd * rsd)
    bmean <- 0.251 / 400
    bsd <- 1.341 * bmean
    btau <- 1 / (0.02)

    a ~ dnorm(amean, atau)T(0,)
    r ~ dnorm(rmean, rtau)T(0,)
    b ~ dnorm(bmean, btau)T(0,)

    tau ~ dgamma(1,0.001)

    #--------

    # model

    # DO modelled
    zz ~ dnorm(0,1000000)
    DO_mod[1] <- DO_obs[1] + zz # make monitorable

    for (i in 1:(num.measurements-1)){

      DO_obs[i+1] ~ dnorm(DO_mod[i+1], tau)
      DO_mod[i+1] <- DO_mod[i]
        + a * PAR[i]
        - r
        - (1 / H[i]) * (-b * pow(U10[i], 2) * pow(sc[i] / 600, -0.5) * (DO_sat[i] - DO_mod[i]))

      # posterior predictive assessment nodes #
      ats[i] <- a
      bts[i] <- b
      gppts[i] <- a * PAR[i]
      erts[i] <- r
      gets[i] <- (1 / H[i]) * (b * pow(U10[i], 2) * pow(sc[i] / 600, -0.5) * (DO_sat[i] - DO_mod[i]))

    }

}