Coupmodel Coupled heat and mass transfer model for soil-plant-atmosphere systems Lorenz Walthert and Elisabeth Graf-Pannatier Swiss Federal Institute for Forest, Snow and Landscape Research (WSL)
Coupmodel - some general information What is Coupmodel coupled heat and mass transfer model consists of different coupled sub-models has been developed for modeling hydrological or thermal processes in the soil-plantatmosphere systems Central part of Coupmodel are two coupled differential equations for water and heat flow, that consider the law of conservation of mass and energy and Darcy's law (water flows occur as a result of gradients in water potential) and Fouriers's law (energy flows occur as a result of gradients in temperature) The soil is integrated in Coupmodel as soil profile with one ore several layers Single- or multirun-mode is possible The model has been developed by Per-Erik Jansson et al. (1979-2010) Coupmodel is available on the Internet (free of charge): http://www.lwr.kth.se/vara%20datorprogram/coupmodel/index.htm
Coupmodel - data input I The calculations of water and heat fluxes are based on soil properties such as the water retention curve functions for the unsaturated and the saturated hydraulic conductivity missing hydraulic soil properties can be derived from a database, that is integrated in Coupmodel (Coup-internal PTF) Important plant properties are how the plants regulate water uptake from the soil and transpiration how the plant cover influences both aerodynamic conditions in the atmosphere and the radiation balance at the soil surface the development of leaf area in the course of the year Meteorological data are the driving variables to the model (time series) most important are precipitation and air temperature air humidity, wind speed and cloudiness are also of interest
Coupmodel - data input II How can you enter data into the Coupmodel? Switch: is a tool to define the configurations of Coupmodel for a given simulation Switches are grouped in 21 modules each with several options and values e.g.: Module Option Values Soil Hydraulic Hydraulic Functions VanGenuchten/Brooks&Corey Plant Root Distribution linear/exponential/table Parameter: is a single input constant with a default value and a potential range of values for each parameter. Parameters are grouped in 19 Modules, e.g.: Module Parameter Value Default Min Max Interception WaterCapacityPerLAI (mm/m 2 ) 0.5 0.2 0.05 1.0 Meteorol. Data Reference Height (m) 40 2 1 100 Parameter Table: is a table that includes one or more parameters (constants). Parameter tables are grouped in 9 modules, e.g.: Module Parameter Table Values_Theta_s (%) Values_alpha (kpa -1 ) Values_n (-) Soil Hydraulic Van Genuchten layer 1 65 0.154 1.533 layer 2 52 0.176 1.425 layer n 43 0.233 1.563 Model File: data input with a data file, e.g., Filename.BIN with daily values of temperature and precipitation
Coupmodel - data input III
Matric potential / hpa Coupmodel - data output Some results of a simulation are obtained as time series (e.g. daily values) for all individual layers of the soil, such as soil temperature water content and water potential water uptake by roots Soil profile x, depth 12-20 cm; daily values, 1976-2006, May-October 0 1971 1976 1982 1987 1993 1998-500 -1000-1500 -2000-2500 -3000 Other results are emitted as time series for the whole site, such as snow depth surface runoff drainage
Interception loss (mm) Own experiences with Coupmodel (2007-2010) We model the water balance of selected Swiss forest sites with two approaches, differing in quality and quantity of the available data 50 45 40 35 30 25 20 15 10 5 0 01.08.1999 01.12.1999 01.04.2000 01.08.2000 01.12.2000 01.04.2001 01.08.2001 01.12.2001 01.04.2002 interception loss (bulk precipitation - throughfall) 01.08.2002 01.12.2002 01.04.2003 01.08.2003 01.12.2003 01.04.2004 01.08.2004 01.12.2004 01.04.2005 01.08.2005 01.12.2005 01.04.2006 01.08.2006 01.12.2006 01.04.2007 01.08.2007 01.12.2007 01.04.2008 01.08.2008 01.12.2008 01.04.2009 01.08.2009 01.12.2009 High quality data 5 Level II plots in Switzerland the main goal is to estimate mass balances and drought stress on the Level II plots many measured parameters for parametrisation, calibration and validation of the Coupmodel are available the data time series cover the last 10 years 2 meteo stations, one in the forest and one in the open land, record air temperature, precipitation, radiation, relative air humidity and wind speed matric potential is measured biweekly with tensiometers equipped with suction cups (water content only since 2009) throughfall is measured biweekly many other measured parameters like soil matrix data or leaf area index are available for the calibration and validation of Coupmodel, we use time series of matric potential and interception Elisabeth Graf-Pannatier is responsible for this high end approach Standard quality data 1000 Swiss forest sites the main goal is to assess the water availability for trees and herbs on these 1000 sites we use the same model calibration than on the Level II plots no measured data of time series are available available are the data of a soil profile and the data of a floristic inventory according to Braun-Blanquet on all sites daily meteorological input-data has been modeled with Daymet in a 100 m resolution for the period 1930-2006 (air temperature, precipitation, radiation and relative air humidity) Lorenz Walthert is responsible for this standard approach
Own experiences plausibilisation of output data We try to plausibilize the results of our water balance simulations When we started to work with Coupmodel, we soon realized that it is not only necessary to calibrate and to validate the model with our own data but also to plausibilize the results with data from the literature. We plausibilize our results with data from the following literature: Literature-study of L. Walthert (2009, written in German) forest-canopy-conductance (gc_max), stand-transpiration, reduction of transpiration due to drought stress (AT/PT) results from sap-flow-measurement-projects, catchment-studies, and eddy-covariance-plots Van der Salm et al. (2007): Water balances in intensively monitored forest ecosystems in Europe interception, evapotranspiration Richard et al. (1978-1987): Physikalische Eigenschaften von Böden der Schweiz (Lokalformen) saturation level (water level) in hydromorphic soils (gley and pseudogley) Hydrologischer Atlas der Schweiz (1992-2010) Runoff-coefficients (runoff/precipitation) from more than 1000 hydrological catchments of Switzerland
Own experiences Coupmodel configuration Some details of our model configuration (switches) Model structure snowpack: on groundwater flow: on lower boundry: no flow - we control the drainage with a drainage pipe (depth, distance) Soil hydraulic lateral water input: off conductivity function: Mualem hydraulic function: Van Genuchten Water uptake Plant basic equation: pressure head response root distribution: exponential
Own experiences - meteorology Meteorology we compared the modeled daily meteorological parameters (Daymet) with the measured ones on 6 Swiss Level II plots the quality of modeled temperature and precipitation is good the quality of modeled radiation and air humidity is moderate in regions with a strong relief, the maximal modeled radiation is 25 % too low on steep, north facing slopes at the beginning, Coupmodel delivered too high evapotranspiration rates; the model produced reasonable values after we increased the reference height of meteorological data from 2 m (default) to 40 m
Own experiences plants I (LAI) Plants I we calculated the maximal LAI of each of our 1000 forest sites with a regression model, that uses simple site and vegetation parameters as input variables; the regression model is based on data from a combined survey of vegetation and hemispherical photographs on 90 Swiss forest plots (Powerpoint-presentation is available) minimal LAI in deciduous forest during winter is roundly 20 % of LAI_max (literature study) for deciduous and larch forests, the development of LAI during the year was derived from phenological data collected by MeteoSchweiz during the last decades
Own experiences plants II Plants II at the beginning, Coupmodel underestimated the interception rates; the model produced reasonable values after we changed the values of WCB, WC/LAI as well as MaxCover and AreakExp (canopy gaps as a function of LAI) the manipulation of WCB and WC/LAI was based on measured interception-timeseries of 5 Swiss Level II plots and in the case of MaxCover und AreakExp on vegetation data of 1000 Swiss forest sites (canopy structure and LAI) with the default value of 20 mm/s for the canopy conductance (gc_max) Coupmodel overestimates the transpiration rates, plausible transpiration values can be obtained with 10 mm/s for deciduous forests, 9 mm/s for mixed forests and 8 mm/s for coniferous forests (literature study) we fixed the rooting depth in all soils at 150 cm, with exception of soils having a Gr-horizon (gleys) or a R-horizon (bedrock) 150 cm seems reasonable after we evaluated the rooting depth and the rooting intensity in roundly 600 Swiss forest soils of our soil database
Saugspannung / hpa AT/potWaterUptake Own experiences soil I (stone content) Horizon OG UG sand silt clay FE-density Corg Skelett nfk (KA5, ohne Skelett) nfk (KA5, mit Skelett) alpha (KA5) n (KA5) m m % % % g/cm3 % % (cm3/cm3) (cm3/cm3) kpa-1 - Ah 0.0 0.3 14 34 52 0.65 9.0 5 0.37 0.35 0.132 5.233 AC 0.3 0.7 22 24 54 0.90 3.2 63 0.25 0.09 0.330 1.352 C 0.7 1.3 26 38 36 1.00 1.0 88 0.22 0.03 0.252 1.530 C 1.3 1.4 26 38 36 1.00 1.0 88 0.22 0.03 0.252 1.530 C 1.4 1.5 26 38 36 1.00 1.0 88 0.22 0.03 0.252 1.530 C 1.5 1.6 26 38 36 1.00 1.0 88 0.22 0.03 0.252 1.530 C 1.6 1.7 26 38 36 1.00 1.0 88 0.22 0.03 0.252 1.530 Stone content (> 2 mm) stones have an influence on the water balance of soils we wish that the plant available water is reduced proportionally to the stone content of the soil therefore we manipulated the pf-curve by reducing the water contents Theta_s, Theta_r and Theta_pwp proportionally to the stone content but we did not change the shape parameters (alpha, n) of the pf-curve when the plant available water amount (nwsk) (50 to 15'000 hpa) gets very small (<5Vol%) due to the manipulation of the pf-curve, Coupmodel fails to simulate the water balance correctly (matric potential shows a high zero-bias in the horizon just above the zone with the critical nwsk of <5Vol%) -1000 1976-2006, May-October, daily values Matric potential 45-70 cm nwsk > 70 cm = 3 Vol%; Ksat: KA5 1975 1980 1986 1991 1997 2002 1976-2006, May-October, daily values Reduction of transpiration AT/PT nfk > 70 cm = 3 Vol%; Ksat: KA5 1975 1980 1986 1991 1997 2002 1.00-3000 -5000 0.90-7000 -9000-11000 -13000-15000 -17000-19000 -21000-23000 0.80 0.70 0.60-25000 Jahr 0.50 Jahr
Saugspannung / hpa AT/potWaterUptake Own experiences soil II (stone content) Horizon OG UG sand silt clay FE-density Corg Skelett nfk (KA5, ohne Skelett) nfk (KA5, mit Skelett) alpha (KA5) n (KA5) m m % % % g/cm3 % % (cm3/cm3) (cm3/cm3) kpa-1 - Ah 0.0 0.3 14 34 52 0.65 9.0 5 0.37 0.35 0.132 5.233 AC 0.3 0.7 22 24 54 0.90 3.2 63 0.25 0.09 0.330 1.352 C 0.7 1.3 26 38 36 1.00 1.0 88 0.22 0.06 0.252 1.530 C 1.3 1.4 26 38 36 1.00 1.0 88 0.22 0.06 0.252 1.530 C 1.4 1.5 26 38 36 1.00 1.0 88 0.22 0.06 0.252 1.530 C 1.5 1.6 26 38 36 1.00 1.0 88 0.22 0.06 0.252 1.530 C 1.6 1.7 26 38 36 1.00 1.0 88 0.22 0.06 0.252 1.530 Stone content (> 2 mm) many simulations of stone rich soils showed that the zero-bias of matric potential gets tolerable when nwsk is bigger than 6 Vol% in all layers of the soil profile. As a consequence, the amount of stones, that can be considered for the manipulation of the pf-curve, is calculated with respect to both the minimal tolerable nwsk (e.g. 6 Vol%) and the nwsk of the soil without stones. 1976-2006, May-October Matric potential 45-70 cm nwsk > 70 cm = 3 Vol%; Ksat: KA5 1976-2006, May-October Reduction of transpiration AT/PT nfk > 70 cm = 6 Vol%; Ksat: KA5-2500 -5000-7500 -10000-12500 -15000 1975 1980 1986 1991 1997 2002 0 1975 1980 1986 1991 1997 2002 1.00 0.90 0.80 0.70-17500 -20000 0.60-22500 -25000 Jahr 0.50 Jahr
Own experiences soil III Organic layer and drought stress We considered the organic layer in our simulations but we realised, that there is a big incertitude concerning both Ksat and the Van Genuchten Parameters in organic layers (Wösten et al., 1999; Hammel and Kennel, 2001; Schramm et al., 2006). Attention: if you want to quantify drought stress as AT/PT, you should use not PT but potwateruptake in the Coupmodel. Be aware that frozen soil influences AT/PT.
Saugspannung / hpa Saugspannung / hpa Coupmodel - open questions - Ksat different PTF provide very different values for Ksat especially for loose and porose soil layers (e.g. KA5, 2005 or Balland et al., 2008). In some soils, the value of Ksat has an extremely high impact on the drying of the soil. Which value of Ksat might be reasonable for porose, humus rich topsoils with a density between 0.6 and 0.9 g/cm 3? 1975 1980 1986 1991 1997 2002 0 1975 1980 1986 1991 1997 2002 0-5000 -5000-10000 -10000-15000 -15000-20000 -20000-25000 1976-2006, May-October, daily values Matric potential 0-5 cm nfk > 70 cm = 6 Vol% Ksat: KA5 (80 mm/d) Jahr -25000 Jahr 1976-2006, May-October, daily values Matric potential 0-5 cm nfk > 70 cm = 6 Vol% Ksat: 3000 mm/d (Balland et al., 2008: 3075 mm/d)
Coupmodel - open questions II We lost a lot of time by searching a reasonable PTF for the Van Genuchten Parameters (Schramm et al., 2006). Can we expect a better PTF from the Multistep Outflow Experiments of Von Wilpert et al.? Are now better PTF s available for organic layers than formulated by Wösten et al. (1999) or by Hammel and Kennel (2001)? How can we model the aeration / air regime of soils with a Darcy based water balance model? What are the consequences on the air regime, when we manipulate the pf-curve due to stones in the soil (reduction of porosity)? To what extent do stones influence the water conductivity? Quantification of lateral flow in hydromorphic soils? Is there any literature available dealing with surface runoff on forest sites? How can we incorporate relief characteristics (concave, convex) in Darcy based water balance models (loss or supply location)? Can we use hydromorphic symptoms of the soil in water balance models?