NCAR Global Mixed Layer Model Version 1.0 NCAR Mixed Layer Ecosystem Model Version 1.0 Draft Documentation Version 1.0 Please send comments, suggestions, and complaints to Keith Moore at jkmoore@ucar.edu. Introduction This directory contains all code necessary to run the marine ecosystem model on the global mixed layer grid presented in Moore et al. (2001a; 2001b). The model is coded in fortran 77 and there are also a number of IDL routines for displaying output and for generating/modifying some of the input files. The file AREADME gives a brief description of all the files in this directory. Global Mixed Layer Grid The spatial resolution of the global surface mixed layer grid corresponds to the top layer of the NCAR NCOM ocean model at the x3prime resolution. This grid has a longitudinal resolution of 3.6 deg. longitude and a latitudinal resolution which varies from 1-2 deg. latitude with higher resolution near the equator. The grid is 102 by 116 gridpoints, however the two end columns are repeated, the actual globe only covers 100 by 116 gridpoints. The repeated columns are holdovers from the NCOM model where they aid in compuational e fficiency. To avoid integrating over these "extra" columns the first and last columns in the landmask file are set to only land values (the model is only run over ocean grid points). Thus, if for some reason you modify the land mask file be sure to mask out the extra columns. Climatological Forcings The model is driven by a number of monthly climatological forcings with an interpolation between these monthly mean values performed at each timestep in the model. Forcings are described in detail in Moore et al. (2001a). We present a brief overview here. Nutrient concentrations below the mixed layer are estimated using the World Ocean Atlas 1998 data (Conkright et al., 1998) for nitrate, phosphate, and silicate. Sub-surface iron concentrations are based on the WOA98 nitrate data and the regional iron/nitrate ratios estimated by Fung et al. (2000) as modified by Moore et al. (2001a). Climatological mixed layer depths are from Monterey and Levitus, (1997). Surface shortwave radiation from the ISCCP data set (Bishop and Rossow, 1991). In the model it is assumed that PAR is 45% of this short wave total (see bio_subs_x3pV5.f). The model could be easily modified to utilize more recent estimates of surface PAR as input. Monthly sea surface temperature values are taken from the WOA98. Vertical velocity at the base of the mixed layer was taken from a 3D run of the NCAR NCOM ocean model. We generated the surface percent sea ice cover dataset using satellite data from the period October 1998-September 1999 (see Moore et al., 2001a for details). Marine Ecosystem Model The marine ecosystem model used here is described in detail by Moore et al. (2001a; 2001b). It includes three explicit phytoplankton groups: diatoms, a generic small phytoplankton group, and the diazotrophs. The diazotrophs are capable of nitrogen fixation and are modeled largely on the known behaviour of Trichodesmium spp. (see Moore et al. 2001a). In addition, calcification by coccolithophores is parameterized as a variable percentage of production by the small phytoplankton group. There is one zooplankton class which is highly parameterized and meant to represent everything from protists to krill. The maximum grazing rate varies depending on the prey. In the real ocean it is different grazers that consume large diatoms and picoplankton with very different grazing rates, reproductive rates, etc... There are two detrital pools one a non-sinking pool meant to represent largely dissolved organic matter but also accounting for very small non-sinking particulates. The second large detrital pool actively sinks out of the mixed layer. Elements within the detrital pool remineralize at different rates as a function of temperature. Atmospheric inputs of iron and silica from mineral dust deposition are included. A constant percentage of iron and silica within the dust is assumed of 3.5% by weight iron and 30.8% by weight silica. The solubility for silica is set at 7.5% in the code. This is the fraction of the deposited silica which dissolves and enters our silicate nutrient pool. This could be modified, Duce et al., 1991 gave an estimated range of 5-10% solubility for Si. The 30.8% Si by weight is also taken from Duce et al., 1991. The percentage iron which dissolves is an input parameter to model and can modified in the input file. In practice our input files have already been converted to unis of nmolFe and this is then converted to Si flux in the code. The atmospheric fe depositions in the provided input files were derived from the modeling studies of Tegen and Fung (1994; 1995) and Mahowald et al. (1999). Please cite these studies if you use the atmospheric depositions in any manner. Running the Model Before running the model you will have to modify some lines in bio_2d_x3pV5.f to reflect your own local directory structure. The code will have to be compiled using a f77 compiler. There is a makefile (MakefileV5) provided which works with our compiler. The model assumes that all input files are availble in the same directory as the executable, but this could be modified. Currently the model integrates fluxes and outputs at the specified intervals. Scalars are not currently time averaged by the model. In practice, we integrate the model for two years as spinup and then output data from a third year at 48 equally spaced ~weekly intervals. So what is written to the output file is a snapshot of the scalar values at the end of the week and the fluxes summed over that week. An IDL routine is provided which sums the quasi-weekly flux data and averages the scalar data into monthly total flux and monthly mean scalar values (the procedure is called avg_weekly in the file x3p_routines). The mixed layer grid uses an adaptive timestep Runge-Kutta integrator (from Numerical Recipes). The adaptive time-step feature sometimes gets hung up when fluxes are very small (typically high-latitude winter). This can lead to stepsize underflow, underflow and inexact errors. These errors are always associated with the integration of fluxes and not with the computation of the scalars. The rounding errors associated with these negligibly small fluxes can thus safely be ignored. On my Sun Ultrasparc1 machine a global 3 year run where output is saved from the 3rd year takes a couple of days. On the faster mainframe machines it is less than a day. One of the really nice features of the mixed layer grid is that since there is no lateral advection included the model can be run only in a selected region or even only at one grid point. To run the model only at the nine JGOFS sites used in the Moore et al. 2001a paper takes only a few minutes. References Bishop, J.K., Rossow, W.B., 1991. Spatial and temporal variability of global surface solar irradiance. J. Geophys. Res., 96: 16839-16858. Conkright, M.E., et al., 1998. World Ocean Atlas Database 1998 CD-ROM Data Set Documentation, National Oceanographic Data Center, Silver Spring, MD. Duce, R.A., et al., 1991. The atmospheric input of trace species to the world ocean. Global Biogeochem. Cycles, 5: 193-259. Fung, I.Y., Meyn, S.K., Tegen, I., Doney, S.C., John, J.C., Bishop, J.K.B., 2000. Iron supply and demand in the upper ocean. Glob. Biogeochemical Cyc., 14: 281-291. Geider, R.J., MacIntyre, H.L., Kana, T.M., 1998. A dyanamic regulatory model of phytoplankton acclimation to light, nutrients, and temperature. Limnol. Oceanogr., 43: 679-694. Mahowald, N., Kohfeld, K., Hansson, M., Balanski, Y., Harrison, S.P., Prentice, I.C., Schulz, M., Rodhe, H., 1999. Dust sources and deposition during the last glacial maximum and current climate: A comparison of model results with paleodata from ice cores and marine sediments. J. Geophys. Res., 104: 15895-15916. Monterey, G., Levitus, S., 1997. Seasonal variability of Mixed Layer Depth for the World Ocean. NOAA Atlas NESDIS 14, U.S. Government Printing Office, Washington, D.C., pp. 96. Moore, J.K., Doney, S.D., Kleypas, J.A., Glover, D.M., Fung, I.Y., 2001. An intermediate complexity marine ecosystem model for the Global Domain. Deep Sea Res. II, in press. Moore, J.K., Doney, S.D., Glover, D.M., Fung, I.Y., 2001. Iron cycling and nutrient limitation patterns in surface waters of the world ocean. Deep Sea Res. II, in press. Tegen, I., Fung, I.Y., 1994. Modeling mineral dust in the atmosphere: Sources, transport, and optical thickness. J. Geophys. Res., 99: 22897-22914. Tegen, I., Fung, I.Y., 1995. Contribution to the atmospheric mineral aerosal load from land surface modifications. J. Geophys. Res., 100: 18707-18726.