Spatial coverage and inter-annual persistence of cryoconite holes on Canada and Commonwealth glaciers, McMurdo Dry Valleys, Antarctica (2014-2015)


This data package includes measurements pertaining to the spatial distribution and multi-annual persistence of cryoconite holes on Canada and Commonwealth Glaciers in Taylor Valley, Antarctica during two consecutive austral summers (2013-14 and 2014-15). Four circular sampling zones were established on each glacier and multiple measurements of the surface shape and absolute location of all cryoconite holes within the sampling zones were recorded. These measurements can be used to generate spatial maps and analyze the persistence of cryoconite holes from one summer to the next by tracking individual holes and identifying the number lost, gained, and persistent on a multiannual scale. The physical state of each cryoconite hole (liquid-filled, fully frozen, or drained/dry) was also recorded in order to assess the capacity for drained columns to re-initiate downward melting later in the austral summer or during the following year. The glacial surface coverage (GSC) of liquid-filled cryoconite holes can be used in order to assess the contribution of these columns to total glacial melt and drainage. A detailed description for how to use these data in subsequent analyses to spatially map and ‘track’ cryoconite holes over time is available in Water Resources Research: A.Q. Mass and D.M. McKnight (2021) The inter-annual persistence and contribution of cryoconite holes in Taylor Valley, Antarctica to the hydrologic cycle of the McMurdo Dry Valleys under a new climate regime. 

Dataset ID: 


Associated Personnel: 


Short name: 


Data sources: 



Notes for data users: Information such as population density, mean diameter, circularity, and the fraction of glacial surface area covered by cryoconite holes at each sampling site can all be calculated directly from the raw data in this package. However, the transect number, CRYO_ORDER, and PERP_DISTANCE_C are relative, date-dependent notes and should not be compared across different sampling dates. The same cryoconite hole may have a different CRYO_ORDER code the following year due to subtle variation in transect placement and the gain/loss of other cryoconite holes (which would in turn affect measurement order). The location-based data in this package can be used to graphically reconstruct each site into a 2-D map of cryoconite holes on the glacial surface. Due to measurement drift (variation in the absolute orientation of transect lines when sites are remeasured), location-based data should not be compared across sampling dates without running a spatial mapping analysis (such as described by Mass and McKnight 2021) in order to align maps from separate measurement periods into the same absolute positioning with a reasonable degree of statistical confidence. Whole dataset comments: Cryoconite holes with a circular surface area (diametric circularity measured as minor/ major axis ratio >0.8) were recorded using measurements A through E (all in centimeters). Conical point F (cm) was only measured for cryoconite holes with an eccentric or oval surface shape (circularity <0.8). Note that ICE_LID_PRESENT is a numeric code for the absence (0) or presence (1) of ice lids and does not count the specific number of lids in multistratified holes.

Four sampling locations were established on each glacier in mid-January 2014. Sites 1-4 were established and measured on Commonwealth Glacier on 1/20/14, followed by sites 5-8 on Canada Glacier on 1/21/14. Three sites on Commonwealth Glacier (CW-1, CW-2, and CW-4) were located 5-10 m downslope from a randomly chosen subset of pre-existing mass balance stakes used as reference locations throughout the ablation zone. The remaining site on Commonwealth Glacier (CW-3) was located 200 m upslope from a meteorological station. Locations on Canada Glacier were determined by a randomization function for both up/downslope and transverse directions 60-1500 m from a central reference site on the glacier. These randomization factors were used in order to prevent sampling bias towards regions with a high number of holes. The site locations were recorded by a handheld GPS and the center of each sampling zone was marked by a 3-m long bamboo post (the ‘center post’) drilled 0.5 m into the ice. A second bamboo post (the ‘reference post’) was installed 3 m upslope from the center post.

Each site consisted of a circular area with a 3 m radius from the center post, encompassing a 28.27 square meter area of the glacial surface. The sampling zone was subdivided into a series of transect ‘wedges’ and the location, size, and physical state of all cryoconite holes within the transect were recorded. Care was taken not to tread on cryoconite holes or other surface features during site installation and measurements. The first transect was conducted by clipping two weight-reinforced 3 m measuring tapes to the center post, orienting the first towards the upslope reference post, and aligning the second tape one meter away following the outer circumference of the site (see Figure 5 in Mass and McKnight 2021). Consecutive transects were then conducted clockwise in 1-meter intervals around the circumference until the full 28.27 square meter circular area was measured. A metal ice screw was drilled into the surface at the circumferential border of transect nine and served as a secondary reference for transect locations midway through the sampling of each site.

Six measurements were collected for each cryoconite hole in the sampling zone, each represented by an alphabetical code in the dataset. These included (A) the distance from the center post to the closest border of the cryoconite hole, (B) distance from the center post to the farthest border of the hole, (C) the perpendicular distance from the lesser transect tape to the closest edge of the hole, (D) the diameter of the major axis, and (E) the diameter of the minor axis. For holes with an eccentric shape, measurement (F) is the distance between the center post and the conical point angled towards (rather than away from) the lesser transect tape. This is a rough measure of the angular orientation of an oval on the glacial surface. A conical point midway between the start/end boundaries (measurements A and B) denotes an oval with a major axis directly perpendicular to the transect line, while a conical point above/below this midway point indicates a major axis tilted towards/away from the transect line with increasing distance from the center post, respectively. Circularity was calculated as the ratio of the minor:major diameter of the hole. Measurement (F) was not recorded for cryoconite holes with circularity >0.8 since the diameter is relatively equal from all angles. (Measurements A through F may be more easily described with diagrams for spatial reference. Multiple figures are available in the Water Resources Research article by Mass and McKnight (2021) and its associated Supporting Information.)

Holes bisected by a transect line were measured during the first of the two intervals. Additionally, holes were characterized as either liquid-filled, frozen (filled with re-frozen meltwater), or fully drained. Liquid-filled holes were subcategorized as ice-lidded or open to the surface. Fully-drained ‘hollow’ cryoconite columns were included in spatial mapping on order to see if any sediment would initiate new melting later in the summer or during the following year. The seasonal timing for transect sampling was kept relatively consistent in the peak of each summer to accurately represent changes over the course of a full year. Measurements were repeated on January 18 and 20, 2015 on Canada and Commonwealth Glaciers, respectively.


Subscribe to RSS - hydrology