--- title: The EDgE modelling chain layout: layouts/skeleton.mustache breadcrumb: - title: Home path: / - title: the EDgE modelling chain path: /The-EDgE-Modelling-Chain --- {{#markdown}} ![Figure 1. Modelling chain to derive the Sectoral Climate Impact Indicators for the Water Sector](modelling-chain.png) The EDgE modelling chain is divided into three main processing steps: 1. Pre-processing and downscaling 2. Hydrological Modelling 3. Post-processing ## Pre-processing & downscaling The pre-processing phase includes all data preparation necessary to transform the land and climate information, including seasonal forecasts and long-term climate projections from different agencies and operational climate services, into computer files that are understood by the hydrological models. One of the key aspects of the project is to integrate the best available land surface characteristics information used by the European Flood Awareness System, EFAS, into the modelling chain. We also use the freely available E-OBS data as a reference climate driving dataset. **Further info:** The daily precipitation and temperature time series from E-OBS are downscaled from the native 25km to a 5km resolution to match the spatial scale of the EDgE domain. A historical hydrological model run using this dataset produces all the EDgE variables for the period 1981-2015. This reference run provides the statistical basis for the calculation of the SCIIs and the statistical downscaling of seasonal meteorological forcings and climate projections. Meteorological data obtained from the North American Multi-Model Ensemble (NMME) and the Copernicus Seasonal Forecasting System are used for the seasonal forecasting of hydrologic variables. The monthly values from these datasets are first downscaled to the 25km E-OBS resolution; this step also removes biases in the long-term mean and distribution function. In a second step, the temporal resolution is increased to daily values using the disaggregation approach presented in Thober et al. (2015). The obtained daily fields at a 25km resolution are then downscaled to the 5km resolution using the same procedure that is used for the reference dataset E-OBS (i.e., external drift kriging). In a third step, the daily values are disaggregated to 3-hourly values using methods described in Bohn et al. (2013) to provide the representation of the diurnal cycle needed by the hydrological models. Monthly values of precipitation and temperature from climate projections are downscaled using the same procedure. The climate projection data are taken from the 5th phase of the Climate Model Intercomparison Project (CMIP5). ## Hydrological modelling In the second step of the modelling chain, both seasonal forecasts and longer-term climate projections are used to force the three hydrological models mHM, VIC, and Noah-MP to capture modelling uncertainty and to obtain the terrestrial Essential Climate Variables (tECVs) and other climate-related water indicators. In EDgE the aim will be to retain soil moisture (SM), river discharge (Q), snow water equivalent (SWE) and Potential Evapotranspration which are the most directly relevant to the water sector. **Further info:** ### Mesoscale Hydrologic Model (mHM): *Model description and setup* The mHM is a spatially distributed, grid-based mesoscale hydrologic model (mHM; Samaniego et al. 2010, Kumar et al 2013a) that accounts for the following main hydrologic processes: canopy interception, snow accumulation and melt, root zone soil moisture and evapotranspiration, infiltration, surface and subsurface runoff, percolation, baseflow and flood routing. The conceptualization of hydrologic processes in mHM is similar to these of other existing large scale models such as the HBV, the WaterGAP, or the VIC models. mHM uses a novel multiscale parameter regionalization (MPR) scheme to account for the sub-grid variability of fine scale physiographical characteristics (e.g., terrain, soil, vegetation characteristics) that facilitates the model to run efficiently across a range of spatial resolutions and locations other than those used in calibration. The source code of mHM is available freely at the www.ufz.de/mhm. The model has been successfully applied to several river basins in Germany, North America and Europe, and other parts of world (Samaniego et al. 2010, 2013; Kumar et al. 2013a,b; Thober et al. 2015, Rakovec et al., 2016). The model is set up over the EDgE domain at a spatial resolution of 5 km. It utilises high resolution land (sub-)surface properties on terrain, soil, vegetation, and geological characteristics to derive effective parameters using the MPR technology. These static land surface characteristics are based on multiple data sources including EU-DEM (EEA) and GTOPO30-DEM, ISRIC SoilGrids1km, CORINE and GLOBCOVER land cover dataset, and IHME1500 Hydrogeological Map of Europe. These datasets are processed, resampled and mapped on to a common resolution of 500 m. During the historical period of 1950-2014, the model is forced with the daily gridded fields of precipitation, air temperature, and potential evapotranspiration - all derived based on the publicly free [E-OBS](http://www.ecad.eu/download/ensembles/ensembles.php) dataset. ### Noah-MP: *Model description and setup* The land surface model (LSM) Noah-MP calculates fluxes and state variables within the energy and water cycles on the terrestrial land surface. It is the successor of the Noah LSM with the inclusion of multiple process parametrization (hence Noah-MP) and can be used as the land surface scheme for the atmosphere Weather Research and Forecasting Model (WRF). It incorporates a large number of process descriptions with a plenitude of parameters. Processes included are among others a two-stream radiation transfer model considering canopy gaps, a Ball-Berry type stomatal resistance scheme, a physically based three-layer snow model and different runoff generation schemes. It distinguishes between surface energy fluxes and states for the canopy and for the ground. Within the EDgE project, the parameters of Noah-MP are calibrated for selected catchments to guarantee a reliable simulation of tECVs. Only the most sensitive parameters, which have been identified in a previous study (Cuntz et al. 2016), are considered in the calibration because the parameter space of Noah-MP is high-dimensional with a total of 150 parameters. The setup of the Noah-MP LSM over the EDgE modelling domain uses the same static data information employed for the mesoscale Hydrologic Model (mHM) when appropriate. These static data encompasses the ISRIC SoilGrids1km soil database and the CORINE land cover dataset. Both of these are given at a spatial resolution that is higher than 5 km. The predominant soil and land cover class are then taken for each 5 km grid cell, which is in line with the model requirements. The vegetation and soil parameters are obtained from standard parameter files for the STATSGO soil dataset and IGBP-MODIS vegetation classes. Additionally, monthly climatological greenness fractions are derived from the [JRC fapar dataset](http://fapar.jrc.ec.europa.eu/Home.php) ### VIC: *Model description and setup* The Variable Infiltration Capacity (VIC) model (Liang et al., 1994, 1996; Cherkauer et al., 2002) simulates the terrestrial water and energy balances. It distinguishes itself from other land surface schemes through the representation of sub-grid variability in soil storage capacity as a spatial probability distribution, to which surface runoff is derived, and base flow from parameterising a deeper soil moisture zone as a nonlinear recession. Horizontally, VIC represents the land surface by a number of tiled land cover classes. Evapotranspiration is calculated using a Penman-Monteith formulation with adjustments to canopy conductance to account for environmental factors. The subsurface is discretized into multiple soil layers. Movement of moisture between the soil layers is modelled as gravity drainage, with the unsaturated hydraulic conductivity a function of the degree of saturation of the soil. Cold land processes in the form of canopy and ground snow pack storage, seasonally and permanently frozen soils and sub-grid distribution of snow based on elevation banding are represented. Soil temperatures are calculated from the heat diffusion equation and ice content is estimated based on temperatures; infiltration and baseflow are restricted based on the reduced liquid soil moisture capacity. The VIC model has been implemented in applications from catchment to global scales for understanding catchment behaviour, extreme hydrological events, hydrological predictability, and climate change impacts (e.g. Sheffield and Wood, 2008; Clark et al, 2014; Sheffield et al., 2014; Yuan et al., 2015). The VIC model is setup for the EDgE modelling domain similar to the other models. Soil parameter values are derived from the ISRIC SoilGrids1km database and adjusted to be consistent with large-scale calibrated values derived from global scale simulations. Land cover spatial variability and associated leaf area index values are taken from AVHRR satellite observations, which are regridded to 5km. ## Post-processing The third step (post-processing) encompasses the calculation of the Sectoral Climate Impact Indicators (SCIIs). The same SCIIs are calculated for all hydrological models and all forcings (i.e., historical, seasonal forecasts, and climate projections). ## Uncertainty analysis There are inherent uncertainties in the SCIIs because of uncertainties in each link of the modelling chain. For example, the historic reference meteorological data have errors, especially where the density of precipitation gauges is low. The hydrological models also introduce uncertainties because of simplifications in the way that they represent the physical processes or missing processes. The seasonal forecasts and climate projections are uncertain for a number of reasons but especially because of the chaotic nature of the atmosphere and uncertainties in its evolution over time. **Further info:** To address the issue of uncertainty, multiple models or sources of information are used. For the hydrological modelling, the mHM, VIC, and Noah-MP models are used to represent the uncertainty in the choice of model. For the seasonal forecasts, a 4-member multi-model ensemble based on the ECMWF CS3 model and U.S. National Multi-Model Ensemble (NMME) is used. Long-term future climate projections are based on the CMIP5 database, which contains data for more than 40 climate models. The uncertainties in the hydrological simulations, and therefore the SCIIs, are quantified in terms of the multi-model, multi-forecast/projection ensemble spread. The uncertainties can be constrained by evaluating the simulations against observations of streamflow, and bias-correcting the tECVs where appropriate. The uncertainty information is communicated in terms of the ensemble spread as well as the source of the uncertainty at different time scales (e.g. from uncertainty in the climate projections due to natural variability). ## References: * Bohn, T. J. , B., Livneh J. W. Oyler, S. W. Running, B. Nijssen, D. P. Lettenmaier, 2013: Global evaluation of MTCLIM and related algorithms for forcing of ecological and hydrological models. Agricultural and Forest Meteorology, 176 , pp. 38-49. * Clark, E. A., J. Sheffield, M. T. H. van Vliet, B. Nijssen, and D. P. Lettenmaier, 2015: Continental Runoff into the Oceans (1950–2008). J. Hydrometeor, 16, 1502–1520. * Cherkauer, K. A., L. C. Bowling, and D. P. Lettenmaier, 2002: Variable Infiltration Capacity (VIC) cold land process model updates, Global Planet. Change, 38, 151–159. * Cuntz M., J. Mai, L. Samaniego, M. Clark, V. Wulfmeyer, O. Branch, S. Attinger, and S. Thober. 2016. "Hard-coded parameters have a large impact on the hydrologic fluxes of the land surface model Noah-MP", Journal of Geophysical Research - Atmospheres, under review. * Samaniego, L., R. Kumar, and S. Attinger (2010), Multiscale parameter regionalization of a grid-based hydrologic model at the mesoscale, Water Resource Research, 46, W05523, doi:10.1029/2008WR007327 * Kumar, R., B. Livneh, L. Samaniego (2013b), Towards computationally efficient large-scale hydrologic predictions with a multiscale regionalization scheme. Water Resource Research, 49(9), doi:10.1002/wrcr.2043 * Kumar, R., L. Samaniego, and S. Attinger (2013a), Implications of distributed hydrologic model parameterization on water fluxes at multiple scales and locations. Water Resource Research, doi:10.1029/2012WR012195 * Liang, X., D. P. Lettenmaier, E. F. Wood, and S. J. Burges, 1994: A simple hydrologically based model of land surface water and energy fluxes for GCMs, J. Geophys. Res., 99(D7), 14,415–14,428. * Liang, X., E. F. Wood, and D. P. Lettenmaier, 1996: Surface soil moisture parameterization of the VIC-2L model: Evaluation and modifications, Global Planet. Change, 13, 195– 206. * Samaniego, L., R. Kumar, and M. Zink (2013), Implications of Parameter Uncertainty on Soil Moisture Drought Analysis in Germany. Journal of Hydrometeorology, doi:10.1175/JHM-D-12-075.1 * Thober, S., R. Kumar, J. Sheffield, J. Mai, D. Schäfer, L. Samaniego: Seasonal Soil Moisture Drought Prediction over Europe using the North American Multi-Model Ensemble (NMME). Journal of Hydrometeorology, 2015, DOI: 10.1175/JHM-D-15-0053.1 * Rakovec, O., Kumar, R., Mai, J., Cuntz, M., Thober, S., Zink, M., Attinger, S., Schafer, D., Schron, M., Samaniego, L. (2016): Multiscale and multivariate evaluation of water fluxes and states over European river basins, J. Hydrometeorol., 17, 287-307, doi:10.1175/JHM-D-15-0054.1. * Sheffield, J., and E. F. Wood, 2008: Global Trends and Variability in Soil Moisture and Drought Characteristics, 1950–2000, from Observation-Driven Simulations of the Terrestrial Hydrologic Cycle. J. Climate, 21, 432–458. * Sheffield, J., E. F. Wood, N. Chaney, K. Guan, S. Sadri, X. Yuan, L. Olang, A. Amani, A. Ali, and S. Demuth, 2014; A Drought Monitoring and Forecasting System for Sub-Sahara African Water Resources and Food Security. Bull. Am. Met. Soc., 95, 861–882. doi: http://dx.doi.org/10.1175/BAMS-D-12-00124.1 * Yuan, X., J. K. Roundy, E. F. Wood, J. Sheffield, 2015: Seasonal Forecasting of Global Hydrologic Extremes: System Development and Evaluation over GEWEX Basins. Bull. Amer. Meteor. Soc., 96, 1895–1912. doi: http://dx.doi.org/10.1175/BAMS-D-14-00003.1 {{/markdown}}