Predicting acidification recovery at the Hubbard Brook Experimental Forest, New Hampshire: evaluation of four models.

The performance and prediction uncertainty (owing to parameter and structural uncertainties) of four dynamic watershed acidification models (MAGIC, PnET-BGC, SAFE, and VSD) were assessed by systematically applying them to data from the Hubbard Brook Experimental Forest (HBEF), New Hampshire, where long-term records of precipitation and stream chemistry were available. In order to facilitate systematic evaluation, Monte Carlo simulation was used to randomly generate common model input data sets (n = 10,000) from parameter distributions; input data were subsequently translated among models to retain consistency. The model simulations were objectively calibrated against observed data (streamwater: 1963-2004, soil: 1983). The ensemble of calibrated models was used to assess future response of soil and stream chemistry to reduced sulfur deposition at the HBEF. Although both hindcast (1850-1962) and forecast (2005-2100) predictions were qualitatively similar across the four models, the temporal pattern of key indicators of acidification recovery (stream acid neutralizing capacity and soil base saturation) differed substantially. The range in predictions resulted from differences in model structure and their associated posterior parameter distributions. These differences can be accommodated by employing multiple models (ensemble analysis) but have implications for individual model applications.

The influence of N load and harvest intensity on the risk of P limitation in Swedish forest soils.

Nitrogen (N) is often considered to be the major factor limiting tree growth in northern forest ecosystems. An increased N availability, however, increases the demand for other nutrients such as base cations and phosphorous (P) which in turn may change which nutrient is the limiting factor. If P or base cations become limiting, N will start to leach which means a risk of increased eutrophication of surface waters. As many studies focus on base cations, this study instead aims at estimating P budgets on a regional scale for different harvesting scenarios relevant for Swedish conditions. P budget calculations were carried out for 14,550 coniferous sites from the Swedish National Forest Inventory, as weathering+deposition-harvesting-leaching. Three scenarios with different harvest intensities were used: 1) no harvesting, 2) stem harvesting and 3) whole-tree harvesting. The input data were derived from measurements and model results. The P budget estimates indicate that harvesting, especially whole-tree harvesting, result in net losses of P in large parts of Sweden. The highest losses were found in southern Sweden due to high growth rate in this area. In the whole-tree harvesting scenario the losses exceeded 1 kg ha(-1) y(-1) on many sites. N budget calculations on the same sites indicate that N generally accumulates in the whole country and especially in the southern parts. Consequently, the N and P budget calculations indicate that the forests in southern Sweden are in a transition phase from N-to P-limitation to growth. This transition will proceed as long as the accumulation of N continues. These results are important in a sustainable forestry context, as a basis for assessing the risk of future N leaching, and in designing recommendations for abatement strategies of N deposition and for application of wood ash recycling and N fertilization.

Modeling recovery of Swedish ecosystems from acidification.

Dynamic models complement existing time series of observations and static critical load calculations by simulating past and future development of chemistry in forest and lake ecosystems. They are used for dynamic assessment of the acidification and to produce target load functions, that describe what combinations of nitrogen and sulfur emission reductions are needed to achieve a chemical or biological criterion in a given target year. The Swedish approach has been to apply the dynamic acidification models MAGIC, to 133 lakes unaffected by agriculture and SAFE, to 645 productive forest sites. While the long-term goal is to protect 95% of the area, implementation of the Gothenburg protocol will protect approximately 75% of forest soils in the long term. After 2030, recovery will be very slow and involve only a limited geographical area. If there had been no emission reductions after 1980, 87% of the forest area would have unwanted soil status in the long term. In 1990, approximately 17% of all Swedish lakes unaffected by agriculture received an acidifying deposition above critical load. This fraction will decrease to 10% in 2010 after implementation of the Gothenburg protocol. The acidified lakes of Sweden will recover faster than the soils. According to the MAGIC model the median pre-industrial ANC of 107 microeq L(-1) in acid sensitive lakes decreased to about 60 microeq L(-1) at the peak of the acidification (1975-1990) and increases to 80 microeq L(-1) by 2010. Further increases were small, only 2 microeq L(-1) between 2010 and 2040. Protecting 95% of the lakes will require further emission reductions below the Gothenburg protocol levels. More than 7000 lakes are limed regularly in Sweden and it is unlikely that this practice can be discontinued in the near future without adverse effects on lake chemistry and biology.

Parameterization and evaluation of sulfate adsorption in a dynamic soil chemistry model.

Sulfate adsorption was implemented in the dynamic, multi-layer soil chemistry model SAFE. The process is modeled by an isotherm in which sulfate adsorption is considered to be fully reversible and dependent on sulfate concentration as well as pH in soil solution. The isotherm was parameterized by a site-specific series of simple batch experiments at different pH (3.8-5.0) and sulfate concentration (10-260 micromol 1(-1)) levels. Application of the model to the Lake Gardsj6n roof covered site shows that including sulfate adsorption improves the dynamic behavior of the model and sulfate adsorption and desorption delay acidification and recovery of the soil. The modeled adsorbed pool of sulfate at the site reached a maximum level of 700 mmol/m(2) in the late 1980s, well in line with experimental data.