Hydrometeorological data from Marmot Creek Research Basin, Canadian Rockies

Meteorological, snow survey, streamflow, and groundwater data are presented from Marmot Creek Research Basin, Alberta, Canada. The basin is a 9.4 km, alpine-montane forest headwater catchment of the Saskatchewan River Basin that provides vital water supplies to the Prairie Provinces of Canada. It was heavily instrumented, experimented upon and operated by several federal government agencies between 1962 and 1986, during which time its main and sub-basin streams were gauged, automated meteorological stations at multiple elevations were installed, groundwater observation wells were dug and 10 automated, and frequent manual measurements of snow accumulation and ablation and other weather and water variables were made. Over this period, mature evergreen forests were harvested in two sub-basins, leaving large clear-cuts in one basin and a “honeycomb” of small forest clearings in another basin. Whilst meteorological measurements and sub-basin streamflow discharge weirs in the basin were removed in the late 1980s, the federal government maintained the outlet streamflow discharge measurements and a nearby high elevation meteorological station, and the Alberta provincial government maintained 15 observation wells and a nearby fire weather station. Marmot Creek Research Basin was intensively re-instrumented with 12 automated meteorological stations, four sub-basin hydrometric sites and seven snow survey transects starting in 2004 by the University of Saskatchewan Centre for Hydrology. The observations provide detailed information on meteorology, precipitation, soil moisture, snowpack, streamflow, and groundwater during the historical period from 1962 to 1987 and the modern period from 2005 to the present time. These data are ideal for monitoring climate change, developing hydrological 20 process understanding, evaluating process algorithms and hydrological, cryospheric or atmospheric models, and examining the response of basin hydrological cycling to changes in climate, extreme weather, and land cover through hydrological modelling and statistical analyses. The data presented are publicly available from Federated Research Data Repository (https://dx.doi.org/10.20383/101.09).

2008) to acoustic measurements of snow (Kinar and Pomeroy, 2009) as well as early telescope based snow surveys (Kinar and Pomeroy, 2015). Utilizing the Cold Regions Hydrological Modelling platform (CRHM) these advances have been synthesised into a physically based hydrological model of MCRB (Fang et al., 2013), which was used to assess the impact of forest disturbances on basin hydrology , analyse antecedent conditions on flood generation (Fang and  and diagnose rain-on-snow runoff generation for alpine environment during the 2013 flood in MCRB (Pomeroy et al., 5 2016).
This paper includes datasets of meteorological, snow survey, streamflow, and groundwater observations measured in MCRB.
Meteorological datasets include historical observations by the CFS and ECCC and recent measurements by the University of Saskatchewan Centre for Hydrology. Continuous records of streamflow measurements by ECCC and University of Saskatchewan as well as groundwater levels monitored by AEP are also included. The snow survey data presented were 10 conducted in clearings, under forest canopies and on hillslopes at various elevations and are useful for model evaluation and snow process studies. Some of the studies utilising these datasets document the basin resilience to changes in climate, extreme weather, and land cover (Harder et al., 2015), a sensitivity analysis of climate warming on snow processes , and assesses variability of climate and its impact on the hydrological processes (Siemens, 2016). is defined by the Water Survey of Canada stream gauge that was installed in 1962 (Bruce and Clark, 1965). MCRB is composed of three upper sub-basins: Cabin Creek (2.35 km 2 ), Middle Creek (2.94 km 2 ), and Twin Creek (2.79 km 2 ), which converge into the Confluence Sub-basin above the main stream gauge (1.32 km 2 ). Upper Marmot Creek is an upper sub-basin 20 of Middle Creek (1.178 km 2 ) is primarily alpine and is also gauged. Based on a 2008 LiDAR 8m digital elevation model (DEM) (Hopkinson et al., 2012), hypsometric curves were derived for MCRB and its three sub-basins (Fig. 2) forests are mainly Engelmann spruce and lodgepole pine (Pinus contorta var. Latifolia) with trembling aspen (Populus tremuloides) present near the basin outlet (Kirby and Ogilvy, 1969). Alpine larch (Larix lyallii) and short shrubs are present around the treeline at approximately 2016 to 2379 m. Exposed rock surfaces, grasses and talus are present in the highest alpine part of basin (1956 to 2829 m). Physiographic descriptions of these ecozones are shown in Table 1 and they are mapped in basin (Golding and Swanson, 1986). The surficial soils are primarily poorly developed mountain soils consisting of glaciofluvial, surficial till and postglacial colluvium deposits (Beke, 1969). Relatively impermeable bedrock is found at the higher elevations, whilst the rest of basin is covered by a deep layer of coarse and permeable soil allowing for rapid rainfall infiltration to subsurface layers overlying relatively impermeable shale (Jeffrey, 1965). Continental air masses control the weather in the region, which has long and cold winters and cool and wet springs with a late spring/early summer precipitation 5 maximum. Westerly warm and dry Chinook (foehn) winds lead to brief periods when the air temperature exceeds 0 o C during the winter monthsthese events can result in snowpack ablation at lower elevations. Annual precipitation ranges from 600 mm at lower elevations to more than 1100 mm at the higher elevations, of which approximately 70 to 75% occurs as snowfall with the percentage increasing with elevation (Storr, 1967). Mean monthly air temperature ranges from 14 o C observed at 1850 m in July to -10 o C observed at 2450 m in January. Mean air temperatures have increased by 2.3 o C from 1967 to 2013, 10 but there are no trends in precipitation or streamflow (Harder et al., 2015). Upper Clearing Tower and Fisera Ridge; the former started data collection 21 October 2007 and the latter started data collection 13 October 2006. The QC data were generated by applying a quality assurance procedure to remove erroneous data in the 15-20 minute raw data. Table 3 lists the QC thresholds used to remove: 1) measurements outside of defined maximum and minimum ranges; 2) measurements that exceed a rate of change (ROC) limit; 3) constant measurements due to sensor failure. In the QC data, values of -9999 denote the measurements removed from the raw data. In addition, daily QC soil moisture is provided for 11 water years from the Level Forest station and eight water years (WY2006 to WY2013) from the Upper Forest. From 19 October 2012, soil moisture is monitored at a 15-minute interval at Upper Forest and this higher temporal resolution data is 25 included.

Recent modelling data
Hourly modelling data were obtained by averaging the 15-minute QC observations of air temperature (°C), relative humidity (%), wind speed (m s -1 ), incoming solar radiation (W m -2 ), and soil temperature at either 5 cm or 10 cm below ground surface (°C) and by summing the 15-minute QC observation of precipitation (mm). Missing observations of air temperature, relative 30 humidity, wind speed, incoming solar radiation, and soil temperature were filled using either temporal averaging interpolation Earth Syst. Sci. Data Discuss., https://doi.org /10.5194/essd-2018-117  or linear regression to nearby stations. When intervals of missing data were less than three hours, temporal averaging was employed where the observations of the variable three hours before and three hours after the missing interval from the same station were used to calculate the average. When the missing data interval was longer than three hours, linear regressions were developed amongst stations using the raw data, the regressions were ranked based on r 2 value, the regression relationship with the highest r 2 value was selected to fill in the missing data. For missing precipitation, observations from nearby station were 5 used along with seasonal precipitation adjustments for elevation to fill in the missing precipitation. The hourly modelling data are provided for 11 water years from 1 October 2005 to 30 September 2016. As described in the previous section, both Fisera Ridge and Upper Clearing Tower stations were established after WY2006, and the hourly modelling data before station establishment were estimated. For the Fisera Ridge station, air temperature, relative humidity, wind speed, incoming solar

Air temperature and relative humidity
Air temperature and relative humidity were measured using Vaisala hygrothermometers at all seven stations.  Fig. 9a and b.

Wind speed
Wind speeds were measured at all seven stations using propeller-type RM Young anemometers. The 11-water year average 30 wind speeds on wind-exposed alpine ridges are 5.8 m s -1 and 2.5 m s -1 at Centennial Ridge measured at 2.41 m a.g.s. (7 m a.g.s.) has an 11-water year average wind speed of 2.0 m s -1 . Vista View station (4.11 m a.g.s.) is located in a large forest cut block with a short sparse forest cover of young trees and has an 11-water year average wind speed of 1.1 m s -1 . For the wind-sheltered stations (Upper Clearing measured at 2.85 m a.g.s, Upper Forest measured at 2.77 m a.g.s, and Level Forest measured at 2.45 m a.g.s), the 11-water year average wind speeds range from 0.1 to 0.6 m s -1 . The maximum hourly wind speed recorded during 11 water years is 37.9 m s -1 from Centennial Ridge station. An example of hourly wind speed from 5 Fisera Ridge station is shown in Fig. 9c.

Incoming solar radiation
Incoming solar radiation was measured at all seven stations using Kipp and Zonen pyranometers and is included in the hourly modelling dataset except for the Vista View station due to the length of measurement. For the Upper Clearing site, hourly incoming solar radiation measured at the top of the 20m tower station is provided in addition to that from the main tripod 10 station near the ground (1.95 m). For the sub-canopy measurements at Upper Forest (i.e. mature spruce forest) and Level Forest (i.e. mature lodgepole forest) stations, the 11-water year mean values range from 15.9 W m -2 (Upper Forest) to 23.7 W m -2 (Level Forest). For the stations not affected by forest canopy, the 11-water year mean value ranges from 140.1 W m -2 (Upper Clearing 20m tower) to 150.3 W m -2 (Fisera Ridge). An example of hourly incoming solar radiation from the Fisera Ridge station is shown in Fig. 9d. 15

Soil temperature
Soil temperature was measured using thermistors at all seven stations at either 5 cm or 10 cm below ground surface. The 11water year mean value ranges from -0.7 °C (Centennial Ridge) to 6.5 °C (Hay Meadow). The maximum hourly soil temperature during 11 water years was 36.6 °C at the Hay Meadow station and the minimum hourly soil temperature during 11 water years was -16.5 °C at the Centennial Ridge station. An example of hourly soil temperature from Fisera Ridge station 20 is shown in Fig. 9e.

Precipitation was measured with Alter-shielded Geonor T200B weighing precipitation gauges at Hay Meadow, Upper
Clearing, and Fisera Ridge stations, and it was corrected for wind-induced undercatch for the wind-exposed Fisera Ridge and Hay Meadow stations (Smith, 2007). Precipitation is divided into rainfall and snowfall based on the psychrometric energy 25 balance precipitation phase determination method developed by Harder and Pomeroy (2013). Table 4 shows that the average annual precipitation for the 11 water years is 627 mm (229 mm snow), 839 mm (443 mm snow), and 1190 mm (802 mm snow) for Hay Meadow, Upper Clearing, and Fisera Ridge, respectively. The highest annual precipitation during the 11 water years from Fisera Ridge station was 1329 mm in WY2013 when approximately 250 mm of rainfall and snowfall fell during the June 2013 flood , which also produced the highest annual rainfall (535 mm) recorded during the 11 water 30 Earth Syst. Sci. Data Discuss., https://doi.org /10.5194/essd-2018-117  years. An example of hourly cumulative precipitation, divided into rainfall and snowfall from Fisera Ridge station, is shown in Fig. 9f.

Historical modelling data
Historical meteorological data is available from the three sites shown in Fig provided. These sites were established in early 1960s by the CFS and ECCC. Based on the availability of data, continuous records of hourly air temperature (°C), relative humidity (%), and wind speed (m s -1 ) and daily precipitation (mm) are included for 18 water years from 1 October 1969 to 30 September 1987. Air temperature and relative humidity were measured by thermographs or hygrothermographs (Munn and Storr, 1967); wind speed was measured by MSC type 45B anemometer, and for precipitation, Leupold-Stevens Q12M weighing gauges and MSC (Meteorological Service of Canada) tipping bucket 10 gauges were used to take measurements for snowfall and rainfall, respectively (Storr, 1973). Data quality assurance was undertaken to generate the continuous data from the original observations, which includes removing inconsistent measurement and outliers, filling missing data with linear regressions to nearby stations. Details regarding the quality assurance are provided by Siemens (2016). The original measured data are also provided for these sites.

Historical snow survey data
Snow survey data collected by CFS from seven snow courses (SC): 1, 3, 6, 8, 11, 14, and 19 are provided for the waters years from 1963 to 1986. The location of these snow courses is shown in Fig. 1, and a brief description for each snow course is listed in Table 5. Regular measurements were carried out monthly from February to June, and each course consisted of 10 staked points where snow depth and snow water equivalent were measured. In some years, measurements were conducted 20 more than once per month, which provided more details of seasonal snow accumulation. Both monthly snow survey data from 1963 to 1986 and detailed survey data from 1963 to 1980 are included for the historical period. 2007 when no measurements were taken. The snow survey data includes snow depth, density and snow water equivalent (SWE). The snow surveys usually occur monthly during the winter accumulation period and bi-weekly to weekly during the spring melt period. Snow depth was measured by a 1-m ruler for shallow snowpack or a 3-m measuring probe for deep snowpacks, and snow density was measured using an ESC30 snow tube for shallow snowpacks or a Mount Rose snow sampler for deeper snowpacks. At each transect, snow depth was observed at 5-m intervals, and one snow density was collected for 30 continuously monitored by AEP. The location of these groundwater wells is shown in Fig. 1, and brief information regarding these wells is provided in Table 6. Daily data for these groundwater wells can be searched and accessed from AEP's "Groundwater Observation Well Network (GOWN)" website at http://environment.alberta.ca/apps/GOWN/. Access to the 5 hourly groundwater well data can be requested from the Groundwater Information Centre at gwinfo@gov.ab.ca.

Example data
Data from the June 2013 flood is shown as an example of weather and streamflow observed in MCRB (Fig. 11). The flood event started on 18 June and ended on 24 June. Air temperature observed at Fisera Ridge station was as high as 8 °C during rainfall on 19 June and dropped to 0.4 °C during snowfall on 21 June; the atmosphere became saturated on 18 June and stayed 10 saturated through 21 June (Fig. 11a). Variable wind speeds were observed at the Fisera Ridge station, changing from relatively calm conditions on 18 June to 4 m s -1 on 20 June then dropping to an average of 2 m s -1 before peaking at 5.5 m s -1 on 21 June (Fig. 11b). Overcast skies persisted during much of the flood event and incoming solar radiation observed at Fisera Ridge station dropped from a peak of 533 W m -2 on 18 June to below 266 W m -2 throughout the event and then rose to a peak of 1038 W m -2 on 22 June (Fig. 11b). Similar depths of precipitation fell at all elevations (1436 to 2325 m) in MCRB, with about 257 15 mm during 19-25 June; however, this measurement was compromised as the Geonor precipitation gauge overtopped on 21 June and could not be immediately accessed for maintenance due to damaged trails and roads. During the snowfall of 21-22 June, the depth of fresh snowpack on the ground was used to estimate precipitation based on assumption of a fresh snow density of 100 kg m -3 . Approximately 237 mm of rainfall was measured at Fisera Ridge station during 19-25 June, and an 8-cm deep snowpack developed at Fisera Ridge on 21 June and melted after 22 June (Fig. 11c). Rainfall 20 and snowfall rates during the event remained less than 12 mm h -1 and were higher than 6 mm h -1 only on 19 and 20 June, with cumulative daily totals increasing from 41 mm on 19 June to 113 mm on 20 June, and then dropping to 77 and 18 mm on 21 and 22 June, respectively. The streamflow discharge observed at outlet of Upper Marmot Creek remained below 0.6 mm h -1 at start of the flood event on 19 June and increased steadily on 20 June, reaching a peak of 2.84 mm h -1 at 1:00 on 21 June and then falling to below 1 mm h -1 after 21 June for the remaining of the flood event (Fig. 11d). Total discharge generated at the 25 outlet of Upper Marmot Creek was estimated to be 106 mm during 19-25 June, much of which was the result of rain-on-snow in the alpine and treeline elevations.

Data availability and structure
All data presented in this paper are publically available at the Federated Research Data Repository  Table 7 lists the theses that can be searched and accessed from University of Saskatchewan's "eCommons" website at https://ecommons.usask.ca/.

Compilation of Marmot Creek Memories, Real-time Data and Publications
The Centre for Hydrology held a 50 th Anniversary Workshop for MCRB in February 2013 where many of the original and recent researchers gave presentations on a half-century of scientific research in the basin. The Centre has also compiled 120 15 MCRB publications, and provides real-time observations from many of the current meteorological stations. The workshop presentations, publications and data can be accessed here http://www.usask.ca/hydrology/MarmotBasin.php.

Figure 9:
Example of hourly-averaged forcing data from Fisera Ridge station showing (a) air temperature, (b) relative humidity, (c) wind speed, (d) soil temperature, and (e) rainfall and snowfall for water years starting 1 October. All data are developed from observations except rainfall and snowfall, which are calculated from wind-corrected storage-gauge observations with precipitation phase calculated as per 5 Harder and Pomeroy (2013). 10 Earth Syst. Sci. Data Discuss., https://doi.org/10.5194/essd-2018-117 Figure 10: Example of mean transect snow accumulation (SWE) from (a) alpine and (b) montane forest snow survey transects. The historical SWE for alpine and montane forest is from SC 19 and SC 3 transects, respectively. The recent SWE for alpine and montane forest is from Fisera Ridge above treeline transects and Upper Clearing forest section transects, respectively. 5 Earth Syst. Sci. Data Discuss., https://doi.org/10.5194/essd-2018-117        East sloping lodgepole pine about 9m tall with natural openings 3 Gently south sloping mature spruce, lodgepole pine and alpine 6 Gently northeast sloping mature spruce, lodgepole pine and alpine fir 8 South sloping lodgepole pine about 6m tall 11 Southeast sloping mature spruce, lodgepole pine and alpine fir 14 Northeast sloping mature spruce, lodgepole pine and alpine fir with small natural openings 19 Variable terrains (i.e. north and south slope, flat and gullies) above treeline