Results from the Radiation Snow Characteristics and Albedo

Results from the Radiation, Snow Characteristics and Albedo at Summit (RASCALS) campaign Aku Riihelä, Panu Lahtinen Finnish Meteorological Institute Teemu Hakala Finnish Geodetic Institute IGARSS 2011 – Vancouver, Canada

I. Motivation and campaign introduction A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 2

Surface albedo of snow • A significant driver of the surface radiation budget of the polar regions • Recent studies indicate that snow/ice albedo treatment in climate models will explain much of the inter-model scatter. • Accurate and robust snow albedo descriptions are needed for detailed studies of polar and even global climate. CM SAF Arctic SAL product, 20. 6. -27. 6. 2011 A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 3

How to get snow albedo? • Spot coverage from AWS stations is not enough; regional and global coverage are required. Ø One may opt to snow parameterizations or direct airborne/spaceborne observations. Ø Airborne coverage sparse in both space and time -> satellite remote sensing is the most viable means to have global, cost-effective coverage. • Regardless of this choice, both parameterization and satellite retrieval are estimates of snow albedo. Ø Careful in-situ observations are needed to validate and develop both. EARSe. L Snow & Ice Workshop, Bern, 7. 2. -9. 2. 2011 4

The RASCALS campaign • RAdiation, Snow Characteristics and Albedo at Summit • Perennial snow during polar summer 2010 • Summit station on top of Greenland Ice Sheet (3200 m) • Pristine snow, no vegetation, no topography EARSe. L Snow & Ice Workshop, Bern, 7. 2. -9. 2. 2011 5

The observations • Surface broadband albedo • Hemispherical-directional snow reflectance, spectrally resolved • Spectrally resolved irradiance • Snow physical characteristics • SSA estimate from macrophotographs • Density, temperature profiles • Surface roughness A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 6

II. Applications for data A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 7

Evaluation of snow albedo parameterizations • The recent Gardner-Sharp snow albedo parameterization was constructed from our observations of snow grains (SSA) and the atmospheric state on site. ECMWF and NCEP NG snow albedo schemes also calculated for illustrative purposes. A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 8

Evaluation of snow albedo parameterizations Snow surface roughness effect in albedo; surface hoar manifesting at SZA > 68 deg. A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 9

Evaluation of snow albedo parameterizations

Evaluation of snow albedo parameterizations

Evaluation of snow albedo parameterizations

Evaluation of snow albedo parameterizations • The GS parameterization performed well across all 4 days (RMSE ~ 0. 01 to 0. 02). • The BC concentrations used to tune the fit were in line with previous knowledge (2 ppb) – but the atmospheric pressure correction term is needed for high glacier snow. • The GCM parameterizations also predicted the general level of albedo, although less accurately. A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 13

Direct validation of satellite reflectance/albedo retrievals Spatial and temporal matching of the observations required! Temporal decorrelation possible – but less of an issue for Arctic glacier snow. EARSe. L Snow & Ice Workshop, Bern, 7. 2. -9. 2. 2011 14

The HDRF/BRF library from Summit The clear-sky BRF dataset covers all Sun zenith angles that are applicable for radiation studies or satellite reflectance retrievals at Summit for this period! EARSe. L Snow & Ice Workshop, Bern, 7. 2. -9. 2. 2011 15

Application; direct validation of satellite reflectance/albedo retrievals Sun Sat EARSe. L Snow & Ice Workshop, Bern, 7. 2. -9. 2. 2011 16

Results against AVHRR / CM SAF SAL Polarized FF measurement, half power A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 17

III. Future directions & conclusions A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 18

Next steps • MODIS instantaneous surface reflectance data for the evaluation would be highly valuable (any MODIS staff present? ) • Proceed from broadband reflectance to channelspecific surface reflectance study • FIGIFIGO data covers 300 -2500 nm waveband • Also allows the evaluation of narrow-to-broadband conversion accuracy A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 19

To conclude… • The RASCALS campaign was successfull. • A sizable database of snow albedo and related quantities was gathered from the Greenland Ice Sheet. • The data is now being applied to improve our understanding of both snow albedo parameterizations and remotely sensed snow albedo estimates. • The snow physics-based Gardner-Sharp parameterization works well with arctic perennial snow (and subarctic snow too). • Comparison of directional reflectance measurements on site versus the CM SAF SAL product shows improvement in the SAL satellite product as the algorithm develops; some underestimation still evident. • Next steps will be to bring new satellite data into the study, and also prepare for the release of data to the scientific community once first publications are out. A. Riihelä / RASCALS campaign / IGARSS 2011, Vancouver 20