Process-based modelling: LandscapeDNDC model

LandscapeDNDC is a new model system, which facilitates scaling of ecosystem processes from the site to regional simulation domains. LandscapeDNDC - partly based on the biogeochemical site scale model DNDC - inherits a series of new features with regard to process descriptions, model structure and data I/O functionality.
It is a newly developed ecosystem model including granular functionalities for simulating biogeochemical C & N cycling, plant growth and the water cycle at site and regional scale.
It belongs to the so-called process-based biogeochemical models which simulate the ecosystem C & N cycling and the associated biosphere–atmosphere exchange on the basis of the underlying plant physiological, microbial and physico-chemical processes. The model enables the combined simulation of different ecosystems of different temporal and spatial scales. All calculations are structured in a modular form each representing different ecosystem components/functionalities, like plant growth, water dynamics, microclimate, soil chemistry and microbiology. Human management impacts such as fertilization or tillage are included as well. The modules originate from different stand-alone models or have been recently developed for the in-corporation into LandscapeDNDC. For a detailed description of the model concept and software implementation, see Haas et al. (2012).
For the simulation of forest ecosystems, the PNET-N-DNDC (Stange et al. 2000) - an advanced DNDC version for forest ecosystems - has been implemented and further developed into LandscapeDNDC. This includes the Pnet-N-DNDCtropica for tropical forest systems (Kiese et al. 2002; Kiese and Butterbach-Bahl 2002 and Werner et al. 2007). For the modelling of arable and grassland ecosystems, the DNDC functionalities regarding crop growth processes and agricultural management activities were introduced into LandscapeDNDC in order to allow for simulations of C & N turnover and exchange for agricultural ecosystems including grassland.
For applying DNDC functionality within LandscapeDNDC the module for microclimate, DNDC soil biochemistry and water dynamics need to be combined with the plant growth modules for either forests (PhysiologyPNET) or arable crops (PhysiologyDNDC). Simulating agricultural ecosystems incorporate management related functionalities representing all effects of agricultural management and practice. Forest management like harvesting or thinning are optional.
The model is using daily meteorological data (max. and min. air temperatures, precipitation, radiation) as well as management data (e.g., seeding/harvesting, tillage, fertilization) as drivers and information on soil and vegetation properties (e.g., texture, pH, crop types) as initialization parameters to calculate daily rates of plant N uptake, litter production, mineralisation, nitrification, denitrification etc.. The model’s soilchemistry module explicitly consider nitrification, denitrification as well as chemo-denitrification as processes of N₂O production and consumption in soils.
The NO production/ consumtion processes nitrification, denitrification and chemo-denitrification are based on specific loss/ uptake rates. These rates are influenced by environmental conditions such as soil aeration, inorganic N concentrations, soil temperature, moisture or pH.
The model can be applied on the site scale, as well as for three-dimensional regional simulations. For regional applications LandscapeDNDC integrates all grid cells synchronously forward in time. This allows easy coupling to other spatially distributed models (e.g. for hydrology or atmospheric chemistry) and efficient two-way exchange of states.
The model compiles on Unix/Linux, Apple OSX and Windows using standard C++ compilers. The model as well as the model source code (including all pre- and postprocessing tools) is available upon request.

To demonstrate the capabilities of LandscapeDNDC for C and N turnover at site scale we demonstrate two test cases based on a 5 year simulation of corn, seeded at the 8th of may and harveted at the 28th of september each year. Fretilization consists of one application of 140 kg N / ha of ammonium-nitrate fertilizer. Agricultural practice contains a tillage event before seeding and one event just after harvesting. 

1. A comparison of the effect of no-tillage agricultural management practice on the soil C stocks and N2O emissions / NO3 losses
2. A comparision of the effect of cultivating winter cover crops on the N2O emissions / NO3 losses

Results

a) No-Till effects on soil C stocks and N turnover

No-tillage agricultural practice leads to an increase in soil C stocks in the first soil layers where all the plant litter (residuals) accumulates and mineralizes whereas for tilled agricultural systems, plant litter (c and N) will be distributed within the tillage depth resulting in an additional C and N source in deeper soil layers of the tillage horizon. 

The baseline simulation results in accumulated N2O emissions of 4.53 kg N2O-N/ha and nitrate losses of 9.25 kg NO3-N/ha for the five years. The soil C stocks accumulate for the baseline of 3564.90 kg-C/ha and for the no-tillage simulation of 3827.40 kg-C/ha for the five years which is a significant increase in soil C sequestration, while differences N2O and NO3 losses are also significantly smaller for the no-tillage practice.  

Effect of modelling contrasted management practices in an arable soil ecosystem after 5 years.

Fig of plant growth baseline.

Fig of N2O emission and NO3 losses baseline or no-tillage.

b) Effects of winter cover crops on soil C stocks and  N turnover

The cultivation of winter cover crops have significant influence on the soil C and N turnover as these crops store available nutrients released by litter decomposition after crop harvest in their biomass over winter preventing gaseous and aquatic nutrient losses. 

In this study we demonstrate the effect of cultivating rapeseed and mustrard as winter crops in our 5 year corn rotation. 

The cultivation of rapeseeds as winter crops lead to an significant reduction in N2O emissions and nitrate losses going along an increas in soil C sequestration for the five years corn rotation. Mustard as cover crop lowers the reactive nitrogen losses as well, results in a smaller increase in soil C.

Effect of modelling different crop rotation patterns in an arable soil ecosystem after.

Fig of plant growth for the rapeseed simulation.

Fig of N2O emission and NO3 losses baseline / rapeseed / mustard. 

Contact: Klaus Butterbach-Bahl, Ralf Kiese and Edwin Haas.

<< back to Techniques