The general objective of this paper is to estimate the snow water equivalent(SWE) of the La Grande River watershed (northern Quebec), usingpassive-microwave data from the SSM/I sensor. Particular emphasis is placedon the analysis of SSM/I multitemporal variations.

The analysis of a database containing observations for three winters showsthat the brightness temperatures of the snow decrease as the SWE increasesfor shallow snow covers. However, when the SWE is >180–200 mm, therelationship reverses. This is directly linked to the fraction of large snowcrystals in the snow cover, since these are responsible for most of thevolume scattering. The snow emissivity is lower for shallow snow covers,since the higher temperature gradient is responsible for the quick formationof large snow crystals. For SWE >80–200 mm, the temperature gradientdecreases and large crystal formation is minimal. Since volume scattering islower, snow emissivity tends to increase. The observations confirm what wasobserved by Mätzler and others (1982)and Mätzler (1994).

Two regression lines were used to estimate the SWE for the beginning and theend of winter. This approach appears to be better, since it takes intoaccount the structure of snow cover. The results were used to deriverepresentative maps of the SWE.

Northern Hemisphere snow cover varies greatly through the year, and thepresence of snow has a large impact on interactions between the land surfaceand the atmosphere. This paper outlines the representation of snow cover inthe Hadley Centre GCM, and compares simulated snow cover with satellite andground-based observations. Climate warming in a simulation with increasedconcentrations of CO2 and sulphate aerosols is found to lead tolarger reductions in snow cover over North Ameriea and Europe than overAsia.

Snow cover in the intermittent snow zone of the Sierra Nevada can occupymore than 10 000 km2 of the mountain range, but it has receivedrelatively little attention in river forecasting. Snow is deposited at lowerelevations only during the cold storms of winter, and remains there only fora few days or weeks. When cold storms have created a thin snow cover at lowelevations, a subsequent warm storm can melt this snow in just a few hoursand increase the runoff response dramatically. Operational hydrologicalmodels and river-forecasting procedures have tended to overlookcontributions from the intermittent-snow zone, focusing instead onrainfall-runoff or melt from the snowpack zone at higher elevations.Data-collection efforts are minimal in this zone, too. Ideally, spatiallydistributed models of snowmelt and runoff generation are needed to accountfor the typically large differences in snow cover on different aspects inthe intermittent snow zone. Although aircraft and satellite imagery would bemost desirable to monitor the distribution of snow cover in theintermittent-snow zone, even a few climate stations that reportprecipitation type and snow presence would be a major improvement over thepresent situation in the Sierra Nevada.

The general status of cryospheric datasets required in climate model studiesis reviewed. Datasets are necessary as boundary conditions, and forvalidation purposes. The former application is decreasing as cryosphericvariables are increasingly being derived prognostically. By contrast, thescope of cryospheric parameters that can be validated is expanding.Cryospheric datasets suitable for validation studies are reported and areaswhere data are lacking are identified.

The perennial ice concentration in the Beaufort Sea was examined usingactive- and passive-microwave observations. We compared the ice type andconcentration estimates from SSM/I and ERS-1 SAR data over a seasonal cyclefrom January 1992 to January 1993. It was found the multi-year (MY)ice-concentration estimates from the SAR data were very stable and werenearly equivalent to the ice concentration estimated at the end of theprevious summer. We contrast this with the variability of the MYice-concentration and ice-fraction estimates obtained using the NASA Teamalgorithm. The passive- and active-microwave algorithms provide total iceconcentrations that are comparable during the winter, hut the passiveestimates are significantly lower during the summer. Passive-microwaveestimates of multi-year-ice concentrations are consistently lower (up to30%) than those from the SAR data. We discuss reasons for thesediscrepancies and the possible biases introduced by the active and passivealgorithms.

Although the formation and melt of sea ice are primarily functions of theannual radiation cycle, atmospheric sensible-heat forcing does serve todelay or advance the timing of such events. Additionally, if atmosphericconditions in the Arctic were to vary due to climate change it may havesignificant influence on ice conditions. Therefore, this paper investigatesa methodology to determine melt-onset dale distribution, both spatially andtemporally, in the Arctic Ocean and surrounding sea-ice covered regions.

Melt determination is made by a threshold technique using the spectralsignatures of the horizontal brightness temperatures (19 GHz horizontalchannel minus the 37 GHz horizontal channel) obtained from the SpecialSensor Microwave Imager (SSM/I) passive-microwave sensor. Passive-microwaveobservations are used to identify melt because of the large increase inemissivity that occurs when liquid water is present. Emissivity variationsare observed in the brightness temperatures due to the different scattering,absorption and penetration depths of the snowpack from the availablesatellite channels during melt. Monitoring the variations in the brightnesstemperatures allows the determination of melt-onset dates.

Analysis of daily brightness temperature data allows spatial variations inthe date of the snow inch onset for sea ice to be detected. Since the dataare gridded on a daily basis, a climatology of daily melt-onset dates can beproduced for the Arctic region. From this climatology, progression of meltcan be obtained and compared inter-annually.

A suite of Arctic and Antarctic products is being prepared from AdvancedVery High Resolution Radiometer (AVHRR) and ancillary data as part of NASA’sPolar Pathfinder effort. These products consist of twice-daily griddedfields of clear-sky surface temperature, surface albedo and cloud fraction,as well as daily ice velocities, for 1983–96. The products and theirproduction methodology are summarized here, with examples demonstratingapplications of the Pathfinder products for process studies andmodeling.

A one-dimensional, atmospheric boundary-layer model is coupled to athermodynamic ice model to estimate the surface turbulent fluxes over thicksea ice. The principal forcing parameters in this time-dependent model arethe air temperature, humidity, and wind speed at a specified level (eitherat 2 m or at 850 mb) and the down-welling surface radiative fluxes, The freeparameters are the air temperature, humidity, and wind-speed profiles belowthe specified level, the surface skin temperature and ice-temperatureprofile, and the surface turbulent fluxes. The goal is to determine how wellwe can estimate the turbulent surface heat and momentum fluxes using forcingparameters from atmospheric temperatures and radiative fluxes retrieved Iromthe TlROS-N Operational Vertical Sounder TOVS) data.

Meteorological observations from the Lead Experiment (LeadEx, April 1992)ice camp are used to validate turbulent fluxes computed with the surfaceobservations, and the results are used to compare with estimates based onradiosonde observations or with estimates based on TOVS data. We and thatthe TOVS-based estimates of the stress are significantly more accurate thanthose found with a constant geostrophic drag coefficient, with a rool meansquare error about half as large. This improvement is due to stratificationeffects included in the boundary-layer model. The errors in the sensibleheat flux estimates, however, are large compared Io the small mean valuesobserved during the field experiment.

While the capability of global and regional climate models in reproducingcurrent climate has significantly improved over the past few years, theconfidence in model results for remote regions, or those where complexorography is a dominant feature, is still relatively low. This is, in part,linked to the lack of observational data for model verification andintercomparison purposes.

Glacier and permafrost observations are directly related to past and presentenergy flux patterns at the Earth-atmosphere interface and could be used asa proxy for air temperature and precipitation, particularly of value inremote mountain regions and boreal and Arctic zones where instrumentalclimate records are sparse or non-existent. It is particularly important toverify climate-model performance in these regions, as this is where mostgeneral circulation models (GCMs) predict the greatest changes in airtemperatures in a warmer global climate.

Existing datasets from glacier and permafrost monitoring sites in remote andhigh altitudes are described in this paper; the data could be used inmodel-verification studies, as a means to improving model performance inthese regions.

The effect of the spatial pattern of mass balance on the steady-stategeometry of a glacier is examined using a Vialov -Nye glacier-flow modelbased on non-linear internal deformation of the ice with no basal motion.Surface profiles are predicted using a range of spatial variations of massbalance that include uniform shifts that cause a change in the mean andspatial patterns with zero mean representing changes in balance gradient andcurvature. The corresponding effect on geometry induced by the differentmass-balance patterns are described in terms of the volume and measures ofsurface slope and convexity. The change in glacier volume, slope andconvexity induced by uniform changes in mass balance with non-zero mean aremore than one order of magnitude larger than corresponding changes caused byspatial patterns of similar amplitude, but with zero mean. These resultsshow that the mean mass balance contains most of the mass-balanceinformation relevant to the dynamic changes of a glacier. An importantconsequence is that the memory of a change in climate, which is controlledby the consequent volume change, should be insensitive to the spatialpattern other than how that affects the mean. The spatial pattern of massbalance does induce small changes in the shape of the steady-state profile,which indicates the spatial pattern could affect the short time-scaleresponse characteristics.

A prime input variable to uncoupled ice-sheet models, or for estimating themass budget of present-day ice sheets, is the distribution of net surfacemass balance. In most eases this is extrapolated from relatively few directmeasurements over a limited time period, and parameterised in terms ofcontinentality, surface elevation and other broad-scale indicators. Between1989 and 1995 a series of oversnow traverses around the interior of theLambert Glacier basin gathered a comprehensive set of data on snowaccumulation and surface properties, surface climatology, ice-sheetvelocities, elevations and thicknesses. Above the 2000 m level accumulationaverages were found to be 76 kg ma−2a−1 (σ = 74), much lower than at similar elevations inWilkes Land. The traverse route contains three distinct accumulationregimes: a relatively high accumulation zone along the western side despitehigher average elevations, a very low accumulation zone in the south due tothe effect of inereased continentality and an eastern sector that exhibits arain-shadow effect in predominantly easterly wind fields. Inter-annualvariability is high- with 1993 a colder year, recording only half the longerterm average accumulation over the portion of the route that wasmeasured.

An important component of models of the cryosphere is the calculation ofaccumulation rates over polar ice sheets. As a first-order approximation,many models rely on the assumption that temperature is the main controllingfactor for precipitation. However, compilation of available ice-core data,including a new core from Taylor Dome, East Antarctica, suggests thatprecipitation is significantly decoupled from temperature for a largeproportion of both the Greenland and Antarctic ice sheets. While theestimated glacial-to-interglacial change in temperature does not differgreatly among ice cores from each ice sheet, the estimated change inaccumulation rate varies by more than a factor of 2. A simple vapor-pressureparameterization gives reasonable estimates of accumulation in the ice-sheetinterior, but this is not necessarily the case close to the ice-sheetmargin, where synoptic weather systems are important.

The amount of Arctic sea ice predicted by the Hadley Centre Global CilimateModel (GCM) is evaluated using 15 years of passive-microwave data. While theHadley model reproduces the seasonal cycle reasonably well, itunderestimates the total area of sea ice by more than 3 × 106 km2 for most of the year. In the winter months, most of theunderestimate in ice area results from the prediction of far too little icein Hudson Bay and the Sea of Okhotsk, leading to an excess of up to 0.2 PWheat input to the atmosphere from Hudson Bay alone. The surface-energybudget of Hudson Bay is investigated using a mixture of surface observations(POLES), satellite data (ATSR, SSM/I and ISCCP) and output from the GoddardData Assimilation Office analysis. Flux adjustments of the order of 200 Wm−2, resulting from anomalously high sea-surface temperaturesin the Levitus (1982) climatology,are found to be the cause of the model’s underestimation of sea ice in bothHudson Bay and the Sea of Okhotsk. The fact that flux adjustments based onan inaccurate climatology will produce errors, even if the model physics iscorrect, underlines the need both for improved climatologies and for modelsaccurate enough not to require flux adjustment.

Arctic precipitation as depicted in the National Center for EnvironmentalPrediction (NGEP) and the National Center for Atmospheric Research (NCAR)reanalysis effort is evaluated using 6 hourly model output for the period1986–93 in conjunction with gauge-corrected climatologies. Climatologicalfields from the model agree favorably with observations in terms of generalspatio-temporal patterns, but with some notable differences. In particular,the precipitation maximum over the central Arctic (the region north of 70°N)is depicted in July, one month too early. Values are too low from Augustthrough December, resulting in underestimates of annual precipitation ofabout 40 mm. Despite these shortcomings, the modeled precipitation fieldsappear to be sufficiently realistic to represent a base for blending withother data to provide gridded fields suitable for use in climate studies andsea-ice models.

The surface-energy budget of the Arctic Ocean depends on the distribution ofvarious sea-ice features that form by both mechanical and thermodynamicprocesses. Melt ponds, new ice and open water greatly affect thedetermination of surface albedo. However, even basic measurements of somesurface-feature characteristics, such as areal extent of melt ponds, remainrare.

A method has been developed to assess the areal coverage of melt ponds, newice and open water using video data from the Beaufort and Arctic StormsExperiment (BASE). A downward-looking video camera mounted on the undersideof a Hercules C-130 aircraft provided clear images of the surface. Imagesacquired over multi-year ice on 21 September 1994 were analyzed using aspectral technique to determine the areal coverage of melt ponds, new iceand open water. Statistics from this analysis were then compared to previousfield studies and to the Schramm and others (in press) sea-ice model.

Recent studies by several groups have indicated that the performance ofgeneral circulation models (GCMs) over the ice sheets is severely limited bythe relatively low resolution of the models at the margins, where surfaceslopes are greatest. To provide accurate energy-budget estimates,resolutions of better than 0.5° are desirable, requiring nested or multiplegridding and accurate, high-resolution boundary conditions. Here we presenta new, high-resolution (5 km) digital elevation model for the Antarctic icesheet, derived from radar-altimeter data obtained from the geodetic phase ofthe satellite, ERS-1. These data have been combined with the revisedice-thickness grid reported in Bamber and Huybrechts (1996) to produce a bed- and surface-elevation datasetfor use in regional and global climate and paleo-climaie modellingapplications. The real level of spatial detail in the datasets has beenexamined with the aid of Landsat Thematic Mapper data. Imagery around IceStream D, West Antarctica, shows that the revised ice-thickness grid isaccurately geolocated, and contains valuable fine-scale topographic detailbeyond that available from the cartographic version of the data (Drewry, 1983). The surface topography inthe region of the Ross Ice Shelf has been used to illustrate the level ofdetail in both the vertical and horizontal resolution of (he surfacedataset. Laudsat data has also been used to examine features in thesurface-elevation data. In particular, the location of the grounding zone,for Ice Streams D and E, derived from the two data sources shows goodagreement. The results of this validation underscore the utility of the newdatasets for high-resolution modelling, and highlight the limitations of theFolio maps for such applications.

Understanding the interaction of solar radiation with the ice cover iscritical in determining the heat and mass balance of the Arctic ice pack,and in assessing potential impacts due to climate change. Because of theimportance of the ice-albedo feedback mechanism, information on the surfacestate of the ice cover is needed. Observations of the surface slate of seaice were obtained from helicopter photography missions made during the 1994Arctic Ocean Section cruise. Photographs from one flight, taken during theheight of the melt season (31 July 1994) at 76° N, 172° W, were analyzed indetail. Bare ice covered 82% of the total area, melt ponds 12%, and openwater 6%, There was considerable variability in these area fractions onscales < 1 km2. Sample areas >2 3 km2 gaverepresentative values of ice concentration and pond fraction. Melt pondswere numerous, with a number density of 1800 ponds km-2. The meltponds had a mean area of 62 m2 a median area of 14 m2,and a size distribution that was well lit by a cumulative lognormaldistribution. While leads make up only a small portion of the total area,they are the source of virtually all of the solar energy input to theocean.