4. Datasets

A complete water and energy budget assessment for the basin would require the evaluation of the individual terms in Eqs. (3.1)-(3.4) over a wide range of spatial-temporal scales, certainly a non-trivial task for a vast and remote (and thus poorly-observed) region such as the MRB. In this preliminary study, we will focus on basin-scale water and energy budgets on monthly and longer time-scales. Because of the lack of detailed observations that are required to evaluate the water and energy budgets for the area, we have to by necessity rely heavily on assimilated, modeled and remotely-sensed datasets to evaluate the basin-scale water and energy fluxes.

Since the regional datasets are typically of higher data resolution and are more thoroughly validated for the area than the global datasets, the relative merits of the global datasets in representing the components of the water and energy cycle for this northern basin will be assessed by using the regional datasets as reference. This type of assessments is important as the global datasets are widely used in climate studies for the region because of their extensive areal coverage, long data availability periods and readily accessibility. To facilitate this intercomparison between the budgets evaluated from the global and regional datasets, the budget analysis will be performed over the 5-y period from June, 1997 to May, 2002. The adoption of this analysis period is largely dictated by the maximum overalp of the availability periods for the various datasets used in this study (see Tables 4.1 and 4.2).

The weighted mask used to calculate the basin-average budgets from NCEP and ERA can be found here.

Dataset Resolution Coverage Period
CRCM 51 km 1997 Apr - 2003 Dec
CMC 34 km/ 24 km (after 09/98) 1997 Mar - Current
NCEP-R2 2.5 deg global 1979 Jan - Current
ERA-40 2.5 deg global 1957 Sep - 2002 Aug

Table 4.1 Available (re-)analysis and model datasets for the MAGS WEBS

4.1 The Canadian Regional Climate Model (CRCM)

The primary climate model for use within MAGS is the Canadian Regional Climate Model (CRCM; Caya and Laprise, 1999). The version of the model used in this study utilize the third generation physical parameterization package of the Canadian Centre for Climate Modelling and Analysis (CCCma) which includes Canadian Land Surface Scheme (CLASS, Verseghy, 1991; Verseghy et al., 1993) among other improvements over the previous-generation CCCma physics package.

In this study, the CRCM is run at a horizontal resolution of 51 km (true at 60º N) with 29 levels in the vertical, 10 of which are below 850 hPa. The entire domain (100 X 90 points) is indicated in Fig. 4.1, along with elevation, the outline of the Mackenzie Basin, and a 9 point sponge zone used to nest the model. Note that there are numerous lakes evident in Fig. 4.1, though apart from Lake Superior, lakes are not represented in the CRCM simulations described here.

The simulation was performed in "climate mode" from April, 1997 to almost current. Lateral boundary and initial (atmospheric) conditions are specified from the CMC operational GEM global analysis. As CLASS requires initial values of soil moisture and temperature, which can be problematic given that very few observed estimates for these fields exist, especially over such large, high latitude domains. The strategy employed to address the problem was to begin the simulation on April 1 1997 without snow cover but with soil moisture specified at half of saturation values (based on porosity). As shown in MacKay et al. (2003), this appears to have been appropriate for most of the Mackenzie Basin. Temperature in the first soil layer was initialized to the mean atmospheric temperature near the surface from operational analysis (in which the model was nested) for the first day of the simulation. Initial temperature for the third (deepest) soil layer was taken from the annual average surface temperature of the Climatic Research Unit’s half-degree monthly climate data set (New et al., 2000). The second soil layer was initialized as the average of the first and third layers. Other aspects of the simulation can be found in MacKay et al. (2003).

For more information, visit: The Canadian Regional Climate Modelling Network

The RCM Domain

Figure 4.1 CRCM domain for the WEBS simulation. Elevation (shaded, contour interval 200 m), the Mackenzie basin outline, and the 9 point lateral sponge zone are also indicated (adapted from MacKay et al., 2003).


4.2 Canadian Meteorological Centre (CMC) GEM Regional Analysis

In 1997, The Canadian Meteorological Centre (CMC) implemented a unified forecasting system in which a single model (the Global Environmental Multiscale model, GEM) is used in different configurations (global, regional, local). Detailed description of the GEM model and its implementation at CMC can be found in Cote et al. (1998a,b). The regional analyses and forecasts using the GEM system were performed on a domain which covers the whole North America with an original horizontal resolution of 0.33 degree and 28 vertical levels. The horizontal resolution was increased to 0.22 degrees (~24 km) in September, 1998. In addition to changes in the model resolution, the model package of the operational GEM system has also gone through several major changes during the course of its operation. In particular, the system adopted the ISBA (Interactions Surface-Biosphere-Atmosphere, Noilhan and Planton, 1989) land surface model (to replace the “force-restore” module) to represent its surface processes during September 2001.

For more information, visit: Canadian Meteorological Centre


4.3 National Centers for Environmental Prediction -Global Reanalysis 2 (NCEP-R2)

The global reanalysis from NCEP/NCAR (Kalnay et al. 1996) has been widely used in water and energy budget studies because of its easy access on the World Wide Web. The reanalysis was done with the Global Spectral Model (GSM) of 28 vertical levels and T62 resolution. A number of modifications were made to the GSM for the subsequent NCEP-DOE Reanalysis II (hereafter referred to as R-2, see Kanamitsu et al. 2002; Kistler et al., 2001 and Roads et al., 2002), the global reanalysis we use in this study. Improvements to the original GSM physics package include better parameterization of convection, an improved treatment of the boundary layer by including non-local diffusion, refined cloud and radiation calculations, as well as enhanced treatments of snowcover and soil moisture in the new reanalysis. A more complete account of the modification to the original GSM as well as the overall improvements of R-2 over the original reanalysis can be found in here.An inter-comparison of the water and energy budgets for the CSE basins by using the R-2 dataset can be found in Roads et al. (2002).

For more information, visit: NCEP-DOE Reanalysis 2 archive at NOAA-CIRES Climate Diagnostic Center


4.4 European Centre for Medium-Range Weather Forecasts 40-yr Global Reanalysis (ERA-40)

Another global reanalysis dataset used in this study is the new ERA-40 reanalysis from ECMWF. This product covers the period from 1957 to 2002. The 3D-Var technique was applied using the T159L60 version of the Integrated Forecasting System to produce the analyses every six hours. Apart from the increased model resolution, ERA-40 also improves over ERA-15 by the comprehensive use of satellite data, starting from the early Vertical Temperature Profile Radiometer data in 1972, then later including TOVS, SSM/I, ERS and ATOVS data. Cloud Motion Winds were used from 1979 onwards. A detailed study of the ERA-40 surface water and energy budgets for the Mackenzie basin can be found in Betts et al. (2003).

For more information, visit: ECMWF Re-Analysis ERA-40


4.5 Regional observations

(i) In situ and synoptic observations

There are 4 regular rawinsonde sites within the Mackenzie Basin (Figure 4.2). These are at Fort Nelson BC, Fort Smith AB, Norman Wells NT, and Inuvik NT. The four other sites nearby were also used. Launches are conducted every 12 hours. The archive contains soundings from Inuvik since 1950, and from mid 1955 for the other locations within and around the basin. There are 10 principal stations within the basin augmented by 43 auto stations (Figure 4.2). Hourly cloud fraction information for 7 sites are available from 1950, and hourly precipitation type information is available over the same period. Discharge data is collected at approximately 80 sites in the Mackenzie Basin by the Water Survey of Canada (WSC). The furthest downstream measurements, at Fort McPherson and Arctic Red River are available on a daily time scale from 1973.

The Site Map

Figure 4.2 Locations of rawinsonde sites (red diamonds), automatic (green triangles) and manned (black cross) synoptic stations in the MRB and vicincity.

(ii) CANGRID regional gridded surface climate datasets

Recently the Meteorological Service of Canada (MSC) has produced a gridded, monthly climate data set of precipitation and screen temperature for Canada covering the period 1990-current. This data set, known as CANGRID, is used here for comparison with the model outputs. Briefly, this dataset is based on operational climate station data which have been homogenized and adjusted for all known measurement errors. For example, monthly adjustment factors were derived from regression models to correct inhomogeneities in the temperature series, which are often non-climatic steps due to station alterations including changes in site exposure, location, instrumentation, observer, observing program, or a combination of the above (Vincent and Gullett, 1999). For precipitation, the daily rain and snow measurements were corrected separately for all known inhomogeneities (Mekis and Hogg, 1999). For each rain gauge type, corrections to account for wind undercatch and evaporation were implemented. Gauge specific wetting loss corrections were also applied for each rainfall event. For snowfall, ruler measurements were used throughout the time series, to minimize potential discontinuities introduced by the adoption of Nipher shielded snow gauge measurements in the mid-1960's. The trace adjustment is particularly important in northern regions such as the MRB where precipitation amounts are relatively low and many trace events are recorded. These adjustments, for example, result in an average increase of 20% in precipitation within the Mackenzie Basin, eliminating the well documented low bias of gauge measurements (Louie et al., 2002). The final 50 km resolution gridded product is the sum of a gridded climate normal based on the square grid technique of Solomon et al. (1968), and a gridded climate anomaly (Zhang et al., 2000).

(iii) Regional gridded SSM/I SWE dataset

SSM/I passive microwave satellite data have been used to derive regional snow water equivalent (SWE) in the prairie region of western Canada for many years (Goodison and Walker, 1995). Algorithms used to translate the SSM/I brightness temperatures into values of SWE (in mm) have recently been developed for the boreal forest region in western Canada and are being tested for the Mackenzie Basin as a research investigation within MAGS (Walker and Silis, 2001). For the Mackenzie Basin area, four algorithms are used to derive a weighted SWE value for each grid point based on the fractions of (i) coniferous, (ii) deciduous, (iii) sparse forest and (iv) open (non-vegetated) land cover types within the grid square. The resulting SWE was calculated from December to March and are mapped on a 25x25 km equal-area grid that covers western Canada.

4.6 Global satellite or blended datasets

As mentioned earlier, one of the objectives of MAGS WEBS is assess the qualities of selected global climate datasets in representing components of the MRB water and energy cycle. The global datasets used in this study include satellite cloud and radiative products and blended satellite and model-generated global precipitation datasets.

(i) ISCCP clouds and radiation datasets

Monthly average distribution and properties of total cloudiness and cloud types are available on a 280km equal-area grid in the International Satellite Cloud Climatology Project (ISCCP, Rossow and Schiffer, 1991) D2 dataset and these cloud coverage data were used in the budget intercomparisons. Recently, ISCCP has produced a new 18-year (1983-2000) global radiative flux data product called ISCCP FD (Zhang et al., 2004). The product was created by employing the NASA GISS Global Circulation Model (GCM) radiative transfer code and a collection of global data sets describing the properties of the clouds and the surface every 3 hours. The results include the all-sky and clear-sky, upwelling and downwelling, total shortwave (SW = 0.4 - 5 m wavelength) and total longwave (LW = 5 - 200 m wavelength) radiative fluxes at five levels: surface, 680 mb, 440 mb, 100 mb and top-of-atmosphere. All of these results are archived with a resolution of 3 hours and 280 km (equal-area map equivalent to 2.5 degrees latitude-longitude at the equator).

(ii) The GPCC/GPCP global precipitation climatology

The GEWEX sponsored Global Precipitation Climatology Project (GPCP, Huffman et al., 1997) led by the Global Precipitation Climatology Centre (GPCC) and other institutions provides a global precipitation datasets on a 2.5o grid for the period 1979-current. It includes multi satellite-based precipitation estimates and the gauge-based analysis of the GPCC (Adler et al., 2003). Over the continents, rain gauge networks supply in-situ observations, while satellite data are used over large oceanic surface and a rain gauge-based precipitation analysis for the global land-surface prepared at GPCC is used for bias reduction of the satellite-based results.

(iii) The CMAP global precipitation climatology

The National Centers for Environmental Prediction (NCEP) Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP) dataset consists of monthly averaged precipitation rate values for the time period Jan 1979 to the Sep 2004 presented on a global 2.5x2.5 degree grid (Xie and Arkin, 1997). Values are obtained from 5 kinds of satellite estimates (GPI,OPI,SSM/I scattering, SSM/I emission and MSU). The enhanced file also includes blended NCEP/NCAR Reanalysis Precipitation values. The standard (with out NCEP/NCAR Reanalysis data) version has some missing values.

(iv) NVAP and GVaP global water vapor data

The GEWEX Water Vapor Project (GVaP) was conducted to improve the measurement, modeling, and long-term prediction of water vapour in the atmosphere. One achievement of GVaP is the development of a global water vapour climatology data set through the NASA Water Vapor Project (NVAP) (Randel et. al., 1996) which provides 1° gridded total column water vapour, cloud liquid water and water vapour at three vertical layers (1000-700, 700-500, 500-300 mb) for the period from 1988-1999. The climatology combines data from rawinsondes, the Special Sensor Microwave/Imager (SSM/I) and the Television and Infrared Operational Satellite (TIROS) Operational Vertical Sounder (TOVS).

Characteristics of the observational datasets used in this study are summarized in Table 4.2 and additional information can be obtained by following some of the links therein.

Parameter Source Resolution Coverage Period
Precipitable Water Rawinsondes Sites Various - Current
GVaP 1.0 deg global 1988 - 1999 Dec
Snow SSMI 25 km 1978 Dec - 2003 Mar (Dec - Mar)
Surface Temperature CANGRID 50 km 1950 - 2003 Dec
Atmospheric Enthalpy Rawinsondes Sites Various - Current
Precipitation CANGRID 50 km 1950 - 2003 Dec
CMAP 2.5 deg global 1979 - 2003 Sep
GPCP 2.5 deg global 1979 - 2003 Dec
Discharge WSC Sites 1972 - 2002 Dec (Arctic Red)
Radiative Fluxes ISCCP FD 280 km global 1983 Jul - 2001 Jun
Cloud Amount Surface Obs. Sites Various - Current

Table 4.2 Available observations for some WEBS parameters

For more information, visit:
International Satellite Cloud Climatology Project (ISCCP)
Global Precipitation Climatology Project (GPCP)
Global water Vapor Project (GVaP)
Global Runoff Data Centre (GRDC)