Modelling of vegetation-hydrology interaction in mountainous forest basin over the two past decades

Abstract:

1. Introduction

Vegetal cover and water cycle are closely linked. The water balance controls the distribution and productivity of terrestrial vegetation, and in turn, the vegetal cover type is a key determinant for evapotranspiration, drainage and global runoff. The water cycle is influenced by vegetation through stomatal conductance, root pattern and leaf area, interception and transpiration.

The impact of climate change on water resources and runoff depends on the vegetation response to temperature, atmospheric CO₂ concentration and precipitations variations. For example, the response of vegetation at elevated CO₂concentrations is an increasing carbon uptake by plants, photosynthesis and vegetation productivity.

At the global scale, the variability of annual runoff is influenced by geographical distribution of evergreen or deciduous vegetation, as well as variations of precipitation (Peel et al., 2001). Moreover, a reduction of forest cover leads to a decreasing evapotranspiration and then an increase of surface runoff.

In this study, we investigate and assess a numerical model coupling a dynamic vegetation model with a hydrogeological model in order to reproduce the discharge variability at the watershed scale over many decades. The originality and strength of this approach is due to the fact that the model is applied and tested on a watershed studied since more than 25 years. Indeed, the Strengbach watershed, located in Vosges Mountains (NE of France OHGE Environmental Hydro-Geochemical Observatory ohge.unistra.fr) represents an ideal site since monthly hydrological and climate data are available since 1986. The studied period covers the 1987-2009 years.

2. Model Description

2.1. Coupling between LPJ model and 3D hydrological model

The LPJ model uses monthly mean climate data over 1950-2009 period from measured data on Strengbach watershed and from global database (Harris et al., 2004) to estimate the water and carbon exchanges between atmosphere, vegetation and soil and vegetal cover dynamic. At the end of twenty years of simulation, the vegetal cover is established and stabilized. The monthly superficial and deep runoffs calculated by LPJ model are then used by a 3D hydrological model over 1950-2009 period in order to take account the incorporated and intercepted water fluxes by vegetation.

2.2. LPJ model

The Lund-Potsdam-Jena Dynamic Global Vegetation Model (LPJ) is derived from BIOME family of models, and combines mechanistic representation of terrestrial vegetation dynamics, with water and carbon exchanges between soil and atmosphere (e.g. Kaplan, 2001). LPJ includes vegetal cover structure, dynamic, competition between plant functional type, and soil biogeochemistry. It is able to simulate fast processes as feedback through canopy conductance between photosynthesis and transpiration and slow processes as competition resource, tissue turnover, soil organic matter and litter dynamics. In this model, ten plant functional type (PFT) are represented, eight forest and two herbaceous. Each PFT is assigned bioclimatic limits as minimum coldest-month temperature for survival or maximum coldest-month temperature for establishment. This determines then whether PFT can survive and/or regenerate under climatic conditions prevailing in a particular site at a particular time in the simulation.

The LPJ model is fully described by Sitch et al. (2003) and has been extensively validated against observations in terms of vegetation structure, phenology, carbon fluxes and their seasonal variability (Sitch et al., 2003).

2.3. Hydrological model: MODFLOW

The MODFLOW model is a numerical model based on Darcy's law and mass conservation concept developed by McDonald and Harbaugh (1988). This model solves the three-dimensional groundwater flow equation for a porous medium using a finite-difference method. This model is the most widely used model to simulate stream and groundwater interactions.

First, a steady-state calibration of MODFLOW was conducted for Strengbach basin to develop an optimal parameter set. Then, a transient simulation of the model was used on catchment to determine the discharge at outlet.

3. Study Site

3.1. Description

The small Strengbach watershed (80 ha) is located on the eastern side of the Vosges Mountains (North East of France). The topography ranges from 883 m at the outlet to 1146 m at the catchment top (Viville et al., 1993). The mean slope of Strengbach catchment is 15 degrees. The climate is oceanic mountainous with monthly mean temperature ranging from -2 to 14 Celsius degrees. The mean annual precipitation is 1370 mm over the last three decades. The mean annual runoff is 814 mm with strong inter-annual variations ranging from 494 to 1132 mm/yr. The high flow rates occur during the cold period and the lowest water discharges at the end of the summer season. The forest covers 90% of the watershed area and consist of 80% of spruces and 20% of beeches. Two main soil types characterize the catchment: brown acidic soils on northern slope and ochreous podzolic soils on the southern side.

3.2. Climate Data

Climate data over Strengbach watershed are provided by OHGE (Hydro-geochemistry Observatory of Environment) and CRU-TS 3.1 spatial climate data set with a spatial resolution of 0.5 degree latitude versus 0.5 degree longitude (Harris et al., 2014) for the simulation over Twentieth century. The CRU-TS data set includes monthly air temperature, precipitation and amount of wet days for 1950-1986 period (supplemented by field OHGE measurement for 1987-2009 period) and monthly cloud cover for 1950-2009 period.

Comparison between field measurement and CU-TS data for the 1987-2009 period show: 1) an annual mean temperature overestimated of 4 Celsius degrees and 2) a precipitation underestimated of 39% for the CRU-TS data. Thus, the monthly mean anomalies correction have been calculated over 1987-2009 period and have been applied on CRU-TS climate data over 1950-1986 period. The monthly precipitation, amount of wet days and temperature are calculated by Meteo-France atmospheric model ARPEGE/Climate (scenario A1B; Deque, 2007) over 1950-2009 period.

3.3. Soil and hydrogeological information

Soil information including textural fraction (clay, silt, sand), bulk density and porosity have been measured on Strengbach watershed. The mean composition of the surficial layer (0 to 0.5 m) and deep layer (0.5 to 1.5 m) is 15% of clay, 19% of silt, 66% of sand and 9% of clay, 19% of silt, 72% of sand, respectively. To determine hydrogeological parameters a proton Magnetic Resonance Sounding (MRS) campaign was led in 2013. The MRS method (Legchenko et al., 2002) is able to give orders of magnitude for and permeability distributed at the catchment scale at different investigated depths.

4. Results and conclusion

4.1Vegetation dynamic

The vegetal cover distribution calculated by biospheric model is similar to those observed on Strengabch catchment. LPJ model predicts 67% of boreal evergreen PFT, 28% of boreal summergreen PFT (deciduous forest) and 5% of temperate herbaceous over 1987-2009 period. The simulated mean production of biomass and leaf area index are overestimated by about a factor of 2, demonstrated a vegetal cover well-developed and stabilized after 30 simulation years. The measured biomass of vegetation has decreased from 6.4 t/ha/year to 3.68 t/ha/year (Le Goaster et al., 1991) for 30-year-old forest and 85-year-old forest respectively. These results, the observed spruces needles yellowing and 30% of defoliation are obvious symptoms of decay forest over the last decades. These symptoms were similar to those observed in other European sites (e.g, Ulrich, 1984).

The transpiration (water flux incorporated by root pattern in two layers) and the water flux intercepted by vegetal cover represent 28% and 18% respectively of mean precipitation over 1987-2009 period (Figure 1). The comparison between simulated and measured water fluxes is a difficult task because the amount of data is very small.

The simulated evapotranspiration (sum of canopy interception plus transpiration of trees) is overestimated by a factor of 2.3 compared with the measured value made during summer period of year 1989 (Viville et al., 1993). This disparity can be attributed to the forest decline at the Strengbach watershed. The evapotranspiration of the present day disease and less dense forest is certainly lower than predicted in model for a great forest.

4.2 Hydrogeological model and discharge at watershed outlet

A first steady-state simulation was performed and enabled to adjust hydraulic conductivities from several measured water levels (Figure 2). Then, the previously calculated recharge was input in the model and a porosity adjustment enabled to calculate watershed discharge on the period 1987-2009 (Figure 3). Even though the calculated discharge does not strictly fit the measured discharge, orders of magnitude, especially for low flow periods, are respected.

4.3 Test sensitivity: impact of vegetal cover on discharge evolution

To explore the link between vegetal cover and discharge variations at catchment scale, a sensitivity test is performed assuming the complete removal of the vegetation from Strengbach basin. The climate is held constant (including temperature, cloud cover, precipitations and atmospheric CO₂concentration) compared to the reference simulation. But no PFT is allowed to establish over the watershed. Through this sensitivity test, we explore the impact of the land plants on discharge at the watershed outlet, neglecting the canopy interception and leaf transpiration.

Consistent with previous studies, this test demonstrates that low flow periods are completely missed and the total discharge is about twice overestimated.

Our results show that, over le last two decades, the coupled models combining biospheric and hydrologic model is able to reproduce the pattern of discharge at the outlet of Strengbach watershed. Future research could explore and evaluate the modification of hydrology over the 21^stcentury in response to climate change context. This work was performed in the framework of the EC2CO project.

References

Deque, M. (2007), Frequency of precipitation and temperature extremes over France in an anthropogenic scenario: Model result and statistical correction according to observed value. Global and Planetary Change, 57, 16-26.

Harris, I., Jones, P.D., Osborn, T.J., and Lister, D.H. (2014), Updated high-resolution grids of monthly climatic observations - the CRU TS3.10 Dataset.International Journal of Climatology,34: 623-642. Doi: 10.1002/joc.3711

Kaplan, J.O. (2001). Geophysical Applications of Vegetation Modeling. PhD Thesis, 128 pp. University of Lund.

Legchenko A., Baltassat J-M., Beauce A., Bernard J. (2002) Nuclear magnetic resonance as a geophysical tool for hydrogeologists. Journal of Applied Geophysics. Volume 50, Issues 1-2, 21-46.

Le Goaster, S., Dambrine, E., Ranger, J., 1991. Growth and mineral nutrition of a spruce stand on poor soil. I - Evolution of the biomass and the dynamic incorporation of mineral elements. Acta Oecologica, 12, 771-789.

McDonald, M.G., Harbaugh, A.W., 1988. A Modular Three-dimensional Finite-difference Groung-water Flow Model. Techniques of Water-Resources Investigations of US Geological Survey. Book 6, CH.A1, Reston, Virginia.

Peel, M.C., McMahon, T.A., Finlayson, B.L., Watson, F.G.R., 2001. Identification and explanation of continental differences in the variability of runoff. Journal of Hydrology, 250, 224-240.

Sitch, S., et al. (2003), Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ dynamic global vegetation model. Global Change Biology, 9(2), 161-185, doi:10.1046/j.1365-2486.2003.00569.x.

Ulrich, B., 1984. Effects of Air pollution on Forest Ecosystems and Waters: The Principles Demonstrated at a Case Study in Central Europe. Atmosph. Environ. 18, 621-628.

Viville, D., Biron, P., Granier, A., Dambrine, E., Probst, A., 1993. Interception in a mountainous declining spruce stand in the Strengbach catchment (Vosges, France). Journal of Hydrology, 144, 273-282.