%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% The attached datasets are the result of work under consideration for publishing. The datasets provide metadata, summary statistics and estimates of air temperature cooling and decoupling on the mountain glaciers of the world. This work was funded by the EU Horizon 2020 Marie Sklodowska-Curie Actions Grant 101026058 under the project name 'TEMPEST' (wwww.tempestglacier.com). The following files are described: 1) GLACIER_DECOUPLING_OBSERVATIONS_DATABASE.mat 2) FUTURE_DECOUPLING_ESTIMATES_DATABASE.mat %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 1) GLACIER_DECOUPLING_OBSERVATIONS_DATABASE.mat is a Matlab data structure with the following order: DATABASE - > [GLACIER CODENAME] -> [VARIABLE_LIST] where the glacier codename is a three letter code for a given glacier with a two number year identifier following an underscore, where multiple years on a single glacier exist (e.g. 'ARO_22' for Arolla Glacier in 2022). Within each glacier location sub-structure exists the data of interest. The Variable list is as follows: NAME: Name of Glacier (year if parenthesis if applicable) CODE: Codename of glacier GLACIER_LAT: Mean latitude of glacier location in decimal degrees GLACIER_LON: Mean longitude of glacier location in decimal degrees GLACIER_ELE: Mean elevation of glacier location in metres above sea level UTC: UTC time zone of glacier location DATE_START: Start date of observation period in Matlab datetime format DATE_END: End date of observation period in Matlab datetime format AWS_SUM: Total number of 'AWS' observation points on the glacier AWS_TOTAL_OBS: Total number of valid hourly observations across all stations AWS_NAME: Name of AWS station(s) AWS_LAT: Latitude of AWS location in decimal degrees AWS_LON: Longitude of AWS location in decimal degrees AWS_ELE: Elevation of AWS location in metres above sea level (at the time of observation start) AWS_FPL: The flowpath length of the AWS location in metres. FPL is taken direct from the literature where available, or calculated from Matlab TopoToolBox functions (Schwanghart et al., 2010) where unavailable AWS_SHIELD: A cell array indicating whether the temperature measurements are artificially aspirated or not AWS_SLP: Mean slope of AWS location in degrees, extracted from the ASTER GDEM grid cell AWS_ASP: Mean aspect of AWS location in degrees, extracted from the ASTER GDEM grid cell AWS_DEB: Presence of supraglacial debris at AWS location (1 = yes, 0 = no) AWS_TPI: The topographic position index of the AWS location in metres, where negative values indicate higher surrounding terrain in a 1 km search window AWS_dCEN: Distance of the AWS location to the centreline of the glacier, in metres AWS_dEDGE: Distance of the AWS location to a lateral side of the glacier or the terminus, in metres AWS_RGH: Roughness elements of surrounding DEM pixels at the AWS location, corresponding to 2d standard deviation filter (Riley et. al. 1999) AWS_OFF_ELE: Elevation of Off-glacier AWS used to derive TaAmb (see below), in metres above sea level TA_GLA_MEAN: Mean on-glacier air temperature (TaGla) in °C for the date range stated in DATE_START:DATE_END TA_OFF_MEAN: Mean off-glacier air temperature (TaOFF) in °C for the date range stated in DATE_START:DATE_END TA_AMB_MEAN: Mean ambient air temperature (TaAmb) in °C for the date range stated in DATE_START:DATE_END, derived from the extrapolation of TaOFF using a locally-derived lapse rate TLR_MEAN: Mean air temperature lapse rate in °C m^-1, following the glaciological convention where negative lapse rates imply a decrease of air temperature with increasing elevation BIAS: The mean difference in TA_GLA - TA_AMB in °C. RH_MEAN: Mean relative humidity in % from the off-glacier AWS, estimated for the on-glacier AWS locations given a lapse rate of the dew point temperature, subsequently converted Q_MEAN: Mean specific humidity in g kg^-1 for the AWS locations, following the approach for relative humidity TE_MEAN: Mean equivalent temperature in °C, calculated following Matthews et al (2022) using TaAmb and specific humidity TE_MEAN_UP: An upper estimate for the mean equivalent temperature in °C, given the uncertainty of the TLR. TE_MEAN_LO: A lower estimate for the mean equivalent temperature in °C, given the uncertainty of the TLR. SWIN_MEAN: Mean incoming shortwave radiation at the AWS location in Wm^-2, where available (NaN when not available) LWIN_MEAN: Mean incoming longwave radiation at the AWS location in Wm^-2, where available (NaN when not available) FFera: The mean wind speed derived from ERA5Land, for the pixel over the AWS. Values are in m s^1, but are notably slow (a common issue from ERA5 datasets). WIND_ID: ID of AWS station (column of AWS_NAME) where wind data are available and valid WIND_SPD_MEAN: Mean wind speed in m s^-1 for those AWS indicated by WIND_ID WIND_DIR_MEAN: Angular mean wind direction in degrees for those AWS indicated by WIND_ID WIND_DC: Directional consistency (unitless) of wind direction for those AWS indicated by WIND_ID WIND_UWI: Up-glacier wind index (unitless) of wind direction relative to down-glacier flow direction at the AWS locations indicated by WIND_ID, following Shaw et al. (2023), whereby a value of -1 (+1) suggests that the mean wind direction is flowing precisely down-(up-)glacier SNOW_ALB: Mean snow albedo (unitless) for the AWS location, derived from the closest pixel of the daily MODIS MOD10A1 product, extracted for the date range stated in DATE_START:DATE_END from the NASA APPears website SNOW_SCA: As SNOW_ALB, but for the fractional snow covered area of the given pixel (unitless) AWS_K: The estimated k parameter of decoupling (unitless), derived from the ratio of TaAmb and TaGla for each AWS location, given a bisquare robust regression fit in Matlab AWS_R2: The R-squared value of the fit for k, used to exclude calculations of k with low confidence AWS_K_UNC: The uncertainty estiamte of the derived k value, considering the maximum difference of regression k-estimates with randomly perturbed uncertainties of the TLR (for TaAmb) and TaGla %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 2) FUTURE_DECOUPLING_ESTIMATES_DATABASE.mat is a Matlab data structure with the following order: FUTURE_K -> [VARIABLE_LIST] The variables presented are based upon estimates of decoupling and cooling over mountain glaciers given future evolution of climate from a pessimistic SSP5-8.5 CMIP6 ensemble. Decoupling ('k') and cooling are estimated based upon either no change of glacier geometry ('static') or the evolution of glacier length and thickness based upon the model results of Rounce et al. (2023). The Variable list is as follows: RGI_ID: The ID number from the Randolph Glacier Inventory v.6 (Pfeffer et al., 2014) for each glacier considered in this analysis. Note that this excludes ice sheets, ice caps and large periphery glaciers of Antarctica and Greenland (n = 187412) REGION: A number corresponding to the region of each glacier analysed LAT: The mean latitude of each glacier in decimal degrees LON: The mean longitude of each glacier in decimal degrees ELE: The mean elevation of each glacier in metres above sea level YEARS: The years (2000-2099) considered in the analysis LENGTH: The length change of each glacier and each year (n,y), as calculated from Rounce et al. (2023), in metres TA: The elevation-adjusted mean air temperature in °C for each glacier in each year (n,y), derived from ensemble mean bias-corrected CMIP6 estimates of an SSP5-8.5 scenario Q: The elevation-adjusted mean specific humidity in g kg^-1 for each glacier in each year (n,y), derived from ensemble mean bias-corrected CMIP6 estimates of an SSP5-8.5 scenario FF: The elevation-adjusted mean specific humidity in m s^-1 for each glacier in each year (n,y), derived from ensemble mean bias-corrected CMIP6 estimates of an SSP5-8.5 scenario K_STATIC: The mean decoupling value for each glacier and each year (n,y) calculated from a multi-linear regression model, based upon observations in database 1) above and considering no change in glacier geometry K_DYNAMIC: The mean decoupling value for each glacier and each year (n,y) calculated from a multi-linear regression model, based upon observations in database 1) above and where glacier geometry is updated based upon the model results of Rounce et al. (2023) COOLING_STATIC: Future evolution of glacier cooling in °C per glacier and year (n,y), defined as (TA x K) - TA, where no change in glacier geometry occurs COOLING_DYNAMIC: Future evolution of glacier cooling in °C per glacier and year (n,y), defined as (TA x K) - TA, where glacier geometry is updated based upon the model results of Rounce et al. (2023) %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 3) Make_Plots_GlobalCooling_OBSERVATIONS.m A Matlab script to call in the data structure of 1) and plot Figures 1 & 2 of the paper*. * The zip folders for 'FUNCTIONS' and 'GIS' should be unzipped and placed in the same file locations (unless code is adapted for different directories). %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% 4) Make_Plots_GlobalCooling_ESTIMATES.m A Matlab script to call in the data structure of 2) and plot Figures 3 & 4 of the paper*. * The zip folders for 'FUNCTIONS' and 'GIS' should be unzipped and placed in the same file locations (unless code is adapted for different directories).