SnowPex-2 ERA5Snow-SL and ERA5Snow-ML

1) Land reanalysis set-up

SnowPEX-2 is the second snow products inter-comparison coordinated by ENVEO and ECCC.  It includes snow cover and snow water equivalent products.

There are 4 categories of products:

• Observation-based products include purely Satellite products (e.g. AMSR2 SWE), or combined products (e.g. GlobSnow, CCI snow)
• Reanalysis products (e.g. 20CRv3, ERA5, JRA-55, ERA5Snow and ERA5Snow-ML)
• Model products (e.g. Crocus_ERA5, ERA5_land, ...)

One of the issue in ERA5, is a discontinuity in 2004 due to IMS snow cover data assimilated from 2004 only (Mortimer et al, 2020, Zsoter et al., 2019).

ERA5Snow-SL and ERA5-ML are stand-alone land surface reanalysis of ERA5 using consistent snow observing system (in situ snow depth) through the reanalysis period from 1980 to 2020, with the single and multi-layer snow model.

ERA5Snow reanalyses are produced using  the Stand-Alone surface analysis (Fairbairn et al 2019), adapted to run in past periods (using Dinand's CI updates) and using the ECMWF snow analysis (de Rosnay et al 2014,2015) with the single layer and multilayer snow models (Arduini et al JAMES 2019). To reduce the computing time they are both using the OI soil analysis instead of the SEKF.

The full period 1980-2020 is covered with 41 streams experiments initialised on 1st of July of each year. The list of experiment is given below. They are reading in the ERA5 atmospheric anlaysis. The experiments are all conducted at the same resolution than ERA5 (TL639).

SWE and snow density are extracted on a 0.25 regular gid and converted into NetCDF (script to extract SnowPEX data)

2) Example of the results from SnowPex-SL compared to ERA5

3) Trends

Time series of 12 months running window averaged SWE for ERA5 (blue) and ERA5Snow (cyan) and ERAland (pink). Also included offline ERA5-Land re-run at the ERA5 resolution with single layer (black) and multi-layer snow model (red).

Global and North America: Clear stepwise negative trend in 2004 in ERA5 due to absence/activation of IMS assimilation before/from 2004. ERA5snow and ERA5;and have similar trend but with no data assimilation ERA5land has more snow than ERA5snow which assimilates in situ observations

Europe: not strong impact of IMS activation in 2004, due to well constrained snow analysis by dense in situ observations. Similar trend the three products

References:

 Arduini G., G. Balsamo, E. Dutra,  J. Day, I. Sandu, S. Boussetta, T. Haiden  Impact of a Multi‐Layer Snow Scheme on Near‐Surface Weather Forecasts, Journal of Advances in Modeling Earth Systems n. 12, pp. 4687–4710.

https://doi.org/10.1029/2019MS001725

de Rosnay P., G. Balsamo, C. Albergel J. Muñoz-Sabater and L. Isaksen: Initialisation of land surface variables for Numerical Weather Prediction, Surveys in Geophysics, 35(3), pp 607-621, 2014
doi: 10.1007/s10712-012-9207-x

de Rosnay P., Isaksen L., Dahoui M.: Snow data assimilation at ECMWF, ECMWF Newsletter no 143, article pp 26-31, Spring 2015, doi: 10.21957/lkpxq6x5

Fairbairn, D., P. de Rosnay, and P. Browne: "The new stand-alone surface analysis at ECMWF: Implications for land-atmosphere DA coupling", Journal of Hydrometeorology 20, 2023-2042, 2019 https://doi.org/10.1175/JHM-D-19-0074.1

Mortimer, C., Mudryk, L., Derksen, C., ...Kelly, R., Tedesco, M. Evaluation of long-term Northern Hemisphere snow water equivalent products, Cryosphere, 2020, 14(5), pp. 1579–1594 https://tc.copernicus.org/articles/14/1579/2020/

Zsoter, E.,  H. Cloke,  E. Stephens, P. de Rosnay, J. Muñoz-Sabater, C. Prudhomme and  F. Pappenberger: "How well do operational Numerical Weather Prediction setups represent hydrology?",  Journal of Hydrometeorology, 2019, 20, 1533-1552, https://doi.org/10.1175/JHM-D-18-0086.1

Zsótér, E., C. Prudhomme, S. Harrigan, P. de Rosnay, J. Muñoz-Sabater, E. Stephens, "Trends in the GloFAS-ERA5 river discharge reanalysis", , ECMWF Technical Memorandum, TM-871, 2020, doi: 10.21957/p9jrh0xp