Abstract. Natural wetlands constitute the largest and most uncertain source of methane (CH4) to the atmosphere and a large fraction of them are found in the northern latitudes. These emissions are typically estimated using process (“bottom-up”) or inversion (“top-down”) models. However, estimates from these two types of models are not independent of each other since the top-down estimates usually rely on the a priori estimation of these emissions obtained with process models. Hence, independent spatially explicit validation data are needed. Here we utilize a random forest (RF) machine-learning technique to upscale CH4 eddy covariance flux measurements from 25 sites to estimate CH4 wetland emissions from the northern latitudes (north of 45∘ N). Eddy covariance data from 2005 to 2016 are used for model development. The model is then used to predict emissions during 2013 and 2014. The predictive performance of the RF model is evaluated using a leave-one-site-out cross-validation scheme. The performance (Nash–Sutcliffe model efficiency =0.47) is comparable to previous studies upscaling net ecosystem exchange of carbon dioxide and studies comparing process model output against site-level CH4 emission data. The global distribution of wetlands is one major source of uncertainty for upscaling CH4. Thus, three wetland distribution maps are utilized in the upscaling. Depending on the wetland distribution map, the annual emissions for the northern wetlands yield 32 (22.3–41.2, 95 % confidence interval calculated from a RF model ensemble), 31 (21.4–39.9) or 38 (25.9–49.5) Tg(CH4) yr−1. To further evaluate the uncertainties of the upscaled CH4 flux data products we also compared them against output from two process models (LPX-Bern and WetCHARTs), and methodological issues related to CH4 flux upscaling are discussed. The monthly upscaled CH4 flux data products are available at https://doi.org/10.5281/zenodo.2560163 (Peltola et al., 2019).

No abstract

Abstract. Natural wetlands constitute the largest and most uncertain source of methane (CH4) to the atmosphere and a large fraction of them are in the northern latitudes. These emissions are typically estimated using process (bottom-up) or inversion (top-down) models, yet the two are not independent of each other since the top-down estimates rely on the a priori estimation of these emissions coming from the process models. Hence, independent validation data of the large-scale emissions would be needed. Here we utilize random forest (RF) machine learning technique to upscale CH4 eddy covariance flux measurements from 25 sites to estimate CH4 wetland emissions from the northern latitudes (north of 45 °N) during years 2013 and 2014. The predictive performance of the RF model is evaluated using the leave-one-site-out cross-validation scheme and the performance (Nash-Sutcliffe model efficiency = 0.47) is comparable to previous studies upscaling net ecosystem exchange of carbon dioxide or studies where process models are compared against site-level CH4 emission data. Three wetland maps are utilized in the upscaling and the annual emissions for the northern wetlands yield 31.7 (22.3–41.2, 95 % confidence interval), 30.6 (21.4–39.9) or 37.6 (25.9–49.5) Tg(CH4) yr−1, depending on the map used. To evaluate the uncertainties of the upscaled product it is also compared against two process models (LPX-Bern and WetCHARTs) and methodological issues related to CH4 flux upscaling are discussed. The monthly upscaled CH4 flux data product is available for further usage at: https://doi.org/10.5281/zenodo.2560164.

Abstract. The regional climate model REMO was coupled with the FLake lake model to include an interactive treatment of lakes. Using this new version, the Fenno-Scandinavian climate and lake characteristics were studied in a set of 35-year hindcast simulations. Additionally, sensitivity tests related to the parameterization of snow albedo were conducted. Our results show that overall the new model version improves the representation of the Fenno-Scandinavian climate in terms of 2 m temperature and precipitation, but the downside is that an existing wintertime cold bias in the model is enhanced. The lake surface water temperature, ice depth and ice season length were analyzed in detail for 10 Finnish, 4 Swedish and 2 Russian lakes and 1 Estonian lake. The results show that the model can reproduce these characteristics with reasonably high accuracy. The cold bias during winter causes overestimation of ice layer thickness, for example, at several of the studied lakes, but overall the values from the model are realistic and represent the lake physics well in a long-term simulation. We also analyzed the snow depth on ice from 10 Finnish lakes and vertical temperature profiles from 5 Finnish lakes and the model results are realistic.

Droughts can have an impact on forest functioning and production, and even lead to tree mortality. However, drought is an elusive phenomenon that is difficult to quantify and define universally. In this study, we assessed the performance of a set of indicators that have been used to describe drought conditions in the summer months (June, July, August) over a 30-year period (1981-2010) in Finland. Those indicators include the Standardized Precipitation Index (SPI), the Standardized Precipitation-Evapotranspiration Index (SPEI), the Soil Moisture Index (SMI), and the Soil Moisture Anomaly (SMA). Herein, regional soil moisture was produced by the land surface model JSBACH of the Max Planck Institute for Meteorology Earth System Model (MPI-ESM). Results show that the buffering effect of soil moisture and the associated soil moisture memory can impact on the onset and duration of drought as indicated by the SMI and SMA, while the SPI and SPEI are directly controlled by meteorological conditions. In particular, we investigated whether the SMI, SMA and SPEI are able to indicate the Extreme Drought affecting Forest health (EDF), which we defined according to the extreme drought that caused severe forest damages in Finland in 2006. The EDF thresholds for the aforementioned indicators are suggested, based on the reported statistics of forest damages in Finland in 2006. SMI was found to be the best indicator in capturing the spatial extent of forest damage induced by the extreme drought in 2006. In addition, through the application of the EDF thresholds over the summer months of the 30-year study period, the SPEI and SMA tended to show more frequent EDF events and a higher fraction of influenced area than SMI. This is because the SPEI and SMA are standardized indicators that show the degree of anomalies from statistical means over the aggregation period of climate conditions and soil moisture, respectively. However, in boreal forests in Finland, the high initial soil moisture or existence of peat often prevent the EDFs indicated by the SPEI and SMA to produce very low soil moisture that could be indicated as EDFs by the SMI. Therefore, we consider SMI is more appropriate for indicating EDFs in boreal forests. The selected EDF thresholds for those indicators could be calibrated when there are more forest health observation data available. Furthermore, in the context of future climate scenarios, assessments of EDF risks in northern areas should, in addition to climate data, rely on a land surface model capable of reliable prediction of soil moisture. © 2016 Author(s)