Introduction

Iceland is the second least forested country in Europe after Malta. An island in the north of the Atlantic ocean Iceland covers about 103.000 km². According to the latest FAO report about 0,3% of that is considered as forest. The abundance of forest and previous deforestation has lead to a governmental and independent interest in afforestation. Annually more than 5 million seedlings are planted all around Iceland, most of it on farms by individuals that are supported in their actions by the government. Most tree species used are introduced species like Sitka spruce (picea sitchensis), Logdepole pine (pinus contorta) and Cottonwood polpar (populus tricocarpa) . Native species are also used, they include Downy Birch (betula pubecense), Rowan (Sorbus acuparia) and Trembling poplar (populus tremula).

Research

The project aims at looking at the potential forest growth in Iceland under current climate and how forest growth and production will change under future climate scenarios. One of the main interest is to quantify the forest growth in Iceland and therefore answer the question : where can good forest growth be expected?

This research takes use of three different process based ecosystem models to examine potential forest growth in Iceland. The models used are similar in what processes they include but they differ in how the processes are connected within the models. All model overlap in common processes but each of them adds a new feature not described by the others. The input data for all the models is climate data series.

 

Model description

LPJ-GUESS models the structure, growth and dynamics of a terrestrial ecosystem. It includes processes of photosynthesis, respiration, allocation, growth, carbon/water balance and disturbance. LPJ-GUESS adds the stochastic behaviour of plant functional type competition and therefore describes vegetation dynamics where biome shifts can occur.

 

BIOMASS is the simplest model of those three. It simulates forest growth of pre described tree species. Biomass includes processes of photosynthesis, allocation, carbon/water balance and it has been used previously for black cottonwood plantation in Iceland.

 

TRIPLEX is a hybrid model that includes processes of photosynthesis, allocation, tree growth, decomposition, carbon/water balance and it also adds soil nutrients cycling not included in the other models.

 

The project divides into three parts: current climate, future climate and past climate. But before these tasks can be started all the models need to be parameterized before the actual model run can be performed.

 

Parameterization

First step of the work is to parameterize the models to conditions in Iceland and other factors relevant locally. These factors include tree species in use and soil texture information.

 

Tree species physiological parameters will be derived from literature and databases already available. These parameters are used differently within each model so they need to be harmonized for the models used in the project. This is essential for the purpose of comparison of data between the models.

Soil texture will be adopted from the Icelandic Agricultural University: Soil map of Iceland. Soil data used in the models is quite simple but differs between models so it also needs to be harmonized for comparison purposes.

 

Current climate

To evaluate the potential forest growth and distribution under current climate the models will be fed with historical climate series. These climate series will be provided by Dr. Haraldur Ólafsson and are a result of meteorological climate model he is working with. The climate series include daily mean temperature and daily precipitation from 1961 to 2006. Daily incoming radiation (sunshine) will also be introduced to the models but from other sources. These climate data are produced on a 0.5° grid. The grid divides Iceland into about 130 grid cells. Models calculate forest dynamics independently on each grid cell over Iceland. LPJ-GUESS uses annual atmospheric CO₂ concentrations in its calculations. When all the input data has been prepared the models can be run and the data comparison/validation can begin.

 

-Data validation

Comparison of data output from the models and direct measurements is one way of validating model performance. Outputs from the models will be validated to ground/satellite based observations. Icelandic Forest Research has previously collected data about forest growth in various locations around Iceland. The comparison would be in the form of modelled biomass to measured biomass or mean annual increment.

Another method suggested in this project is comparison with satellite data. Satellite imagery is available from various sources but most likely MODIS data will be used in this research. MODIS produces, among other, data about fAPAR (fraction of Absorbed Photosynthetic Active Radiation) that can be translated as LAI (Leaf Area Index) witch is an output from the models. This satellite data comparison can only be done where forests are present today so it is only relevant on a local basis.

Another source of data for comparison is the Icelandic National Forest Inventory (INFI). INFI is a systematic forest measurement scheme over the whole country. This project is underway and 2/5 of the measurements have been carried out.

This validation will be done independently between outputs form each model used in the study. Differences in model outputs and comparison to measured data can then be analyzed and described. 

 

Future climate

Forecasting potential forest growth in the future is done in a quite similar way as described in the chapter about current climate. Same climate series data will be used in this part. The data about future climate change will be derived from literature and IPCC reports and different climate change scenarios will be looked at. Changes in temperature and precipitation will gradually be added or subtracted from the climate series data over the forecasted time limit of the research. Predicted changes of atmospheric CO₂ concentrations will also be increased over the timeframe of the research.

Results from this part of the project can not be validated. We base our results on the validation of model performance from the current climate. Our assumption is that if the models produce acceptable results for current climate they should also be applicable under future climate scenarios.

 

Past climate

When the models have been established and validate for current climate they should be able to describe previous forest distribution and growth. To examine past forest distribution a paleoclimatic dataset is needed. For our purposes this dataset can not be derived from pollen analysis because of positive feedback mechanisms. We plan to use data from sea coring projects to derive variations in temperature. This derived dataset will then be used to reconstruct vegetation and forest dynamics through out the Holocene. Comparison of model output to pollen data series will provide some means of validating the results.

Bjarki Þór Kjartansson

PhD student in Physical Geography and Ecosystem Analysis

PhD Project - Potential forest growth in Iceland