Strengths and weaknesses of the new CERRA-Land  surface reanalysis at 5.5 km resolution over Europe

<p><span>The Copernicus Climate Change Service of the European Commission </span><span>aims </span><span>to</span><span> produc</span><span>e</span><span> and deliver a </span><span>land </span><span>regional reanalysis for Europe covering the period </span><span>from</span><span> 19</span><span>84 to </span><span>the present</span><span> at </span><span>a </span><span>horizontal </span><span>resolution of </span><span>5.5 km.</span></p><p>The need for precipitation and surface variables at an ever-increasing spatial and temporal resolution is a recurrent demand. These variables allow, among other things, to address water resource management issues and to conduct climate change impact studies. Regional surface reanalyses are a way to reconstruct these variables for past periods covering several years using state-of-the-art models.</p><p><span> CERRA-</span><span>Land </span><span>(Copernicus European Regional Re-Analysis)</span><span> is a </span><span>regional</span><span> land surface reanalysis dataset </span><span>that</span><span> describes the evolution of soil moisture, soil temperature and </span><span>the</span><span> snowpack. </span><span>CERRA-Land</span><span> is the result of a single </span><span>stand alone integration of the SURFEX V8.1 land surface model driven by meteorological forcing from the CERRA atmospheric </span><span>reanalysis</span><span> and an offline </span><span>analysis </span><span>of </span><span>daily accumulated </span><span>surface</span><span> precipitation </span><span>using</span> <span>an </span><span>optimal interpolation between a</span><span>n initial estimate (</span><span>first guess</span><span>)</span> <span>based on CERRA </span><span>predicted</span><span> precipitation</span><span> and in situ rain gauges</span><span>. The </span><span>2-meter </span><span>temperature and relative humidity </span><span>forcing data </span><span>come </span><span>from</span> <span>the </span><span>CERRA </span><span>surface analysis and the</span> <span>shortwave and longwave </span><span>downw</span><span>elling </span><span>radiation, the 10-meter wind speed, and surface pressure come from the CERRA forecast output</span><span>s</span><span>. </span></p><p><span>The CERRA-Land system </span><span>uses the tiling approach where each grid-box of the model is </span><span>divided into three different fraction</span><span>s</span><span>: urban, lake and natur</span><span>al</span><span>. </span><span>For the nature fraction, the soil is discretized into 14 layers to accurately describe the water and energy transfers between the surface and the deep soil.</span></p><p><span>The quality of </span><span>CERRA-</span><span>Land is assessed by compari</span><span>sons</span> <span>to</span><span> ground-based observations, </span><span>such as</span><span> snow depth, </span><span>2-</span><span>meter temperature</span><span>, </span><span>downw</span><span>elling</span><span> shortwave radiation </span><span>and independent rain</span><span>fall</span><span> station</span><span>s</span><span>.</span> <span>A comparison with E</span><span>RA</span><span>5-Land will be done and the add</span><span>ed</span><span> value of th</span><span>e</span><span> regional dataset will be discussed. </span><span>The entire </span><span>CERRA</span><span>-Land dataset from 19</span><span>84</span><span> to present is </span><span>now available in the </span><span>MARS</span><span> database </span><span>at ECMWF</span><span> and will be avai</span><span>lable</span><span> through the </span><span>Copernicus C</span><span>limate </span><span>D</span><span>ata </span><span>S</span><span>tore.</span></p>