In this section, national Australia-wide grids of daily precipitation (P; available from 1900 onwards) generated by the Bureau of Meteorology (Jones et al., 2009) are used; they are 0.05 degree (or ~5 km) grid cell resolution. The Penman formulation is used to calculate daily potential evapotranspiration (PET; a measure of the atmosphere’s ‘drying power’), which is calculated per Donohue et al. (2010), with meteorological data, other than daily average wind speed (McVicar et al., 2008), being provided by the Bureau of Meteorology (Jones et al., 2009). The PET data also have 0.05 degree (or ~5 km) grid cell resolution. The daily PET data (1982 onwards, due to use of satellite based albedo (the colour of the landsurface, defining how much sunlight is reflected) in the radiation calculations) and daily average wind speed data (1975 onwards, when the Bureau of Meteorology network of anemometers become suitable for national assessment) are generated, and made freely available, by CSIRO Land and Water.
The climate is sub-tropical, with the long-term (i.e. 1900 to 2012) subregion-average precipitation being approximately 1100 mm/year (Figure 17). Like much of Australia there is considerable inter-annual variability, with some years receiving high precipitation (e.g. 1963 received 1890 mm/year) and consecutive years of lower than average precipitation (e.g. 1979 to 1983) that indicate drought conditions (Figure 17). This analysis shows temporal variability of a key hydrological variable: precipitation. Climate also exhibits spatial variability and Figure 18(a) shows the 1982 to 2012 annual average precipitation varies spatially over the subregion. In the broader vicinity of the subregion this ranges from 960 to 1400 mm/year; the higher precipitation values are associated with higher elevations (Figure 5). In the subregion over the last 30 years (i.e. 1982 to 2012) the annual average precipitation is 1095 mm/year, with the maximum and minimum being 1196 and 1023 mm/year, respectively. PET in the broader vicinity of the subregion varies from 1400 to 1700 mm/year, and, as expected, the spatial pattern is complementary to precipitation. Areas receiving high amounts of precipitation are usually cooler and cloudier, so the PET values are lower in these parts of the landscape. Within the subregion the 1982 to 2012 annual average PET is 1587 mm/year, and with the maximum and minimum being 1622 and 1485 mm/year, respectively.
Figure 17 Temporal characteristics of annual precipitation for the Gloucester subregion
(a) shows subregion-averaged annual precipitation with smoothed rolling average (orange line) and (b) annual precipitation divergence from the long-term (1900 to 2012) mean. Source data: Jones et al. (2009)
Source data: (i) precipitation from Jones et al. (2009); (ii) PET from Donohue et al. (2010)
Within a year there is a strong seasonal cycle in precipitation (Figure 19). On average, the rainy season extends from December to March, with the winter months (i.e. July to September, inclusive) being the drier part of the year. When monthly P is compared to monthly PET we see that P has a similar magnitude to PET with PET being greater than P for most (not all) months. The Gloucester subregion can be considered as ‘equitant’ (i.e. straddling the water-limit and energy-limit) throughout the year (McVicar et al., 2012b). This suggests that actual evaporation (AET) in the Gloucester subregion is slightly water-limited (defined when the PET/P ratio is greater than 1.0, as opposed to being energy-limited; when PET/P is less than 1.0). Given the high amounts of precipitation (relative for Australian conditions) there will be high levels of AET and associated vegetation growth (see Figure 12 and Figure 13).
Source data: (i) precipitation is from Jones et al. (2009); (ii) PET is from Donohue et al. (2010)
Figure 20 shows average monthly conditions over the last 30-years (i.e. 1982 to 2012), and below this there is temporal variability for precipitation, PET and the climatic factors (primarily air temperature, vapour pressure deficit, net radiation and wind speed) that govern PET. As expected, monthly P experiences greater variability when compared to other climatic factors (Figure 20).
Charts show: (a) precipitation, (b) potential evapotranspiration (PET), (c) maximum temperature (Tmax), (d) minimum temperature (Tmin), (e) vapour pressure deficit (VPD), (f) net radiation (Rn), and (g) wind speed for the Gloucester subregion. The mean (solid line), ± 1 standard deviation (dashed lines) and the minimum to maximum range (blue shaded area) are shown. Values were calculated over the years 1982 to 2012 (inclusive).
Source data: (i) precipitation, Tmax, and Tmin are from Jones et al. (2009), (ii) PET, VPD and Rn are from Donohue et al. (2010), and (iii) wind speed is from McVicar et al. (2008).
Monthly trends of precipitation, PET and all variables driving PET are shown in Figure 21. The monthly trends for precipitation straddle the no trend (i.e. zero mm/month/year) line, whereas PET, even in the face of warming air temperatures is declining. Declining rates of PET are due to declining amounts of net radiation and wind speed (in all months) and vapour pressure deficit (in most months), which together result in a larger change than the PET increases associated solely with increasing air temperature. Similar findings were reported for other areas of south-east of Australia (Donohue et al., 2010; Donohue et al., 2011; McVicar et al., 2012a).
Charts show: (a) precipitation, (b) potential evapotranspiration (PET), (c) maximum temperature (Tmax), (d) minimum temperature (Tmin), (e) vapour pressure deficit (VPD), (f) net radiation (Rn), and (g) wind speed for the Gloucester subregion. The trend (line), ± 1 standard error (blue shaded area) and trend significance (markers) are shown. Values were calculated over the years 1982 to 2012 (inclusive). Trends are obtained from ordinary linear regression (a parametric test) of the monthly time series and significance was calculated using 2-sided T-test (another parametric test).
Source data: (i) precipitation, Tmax, and Tmin are from Jones et al. (2009), (ii) PET, VPD and Rn are from Donohue et al. (2010), and (iii) wind speed is from McVicar et al. (2008).
While future climate projections produced by global climate models are unsure (GCMs; Lim and Roderick, 2009; Sun et al., 2011), one approach is to use their output and assess what future projections of rainfall and runoff will be. Using 15 CGMs from the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC, 2007, hereafter referred to as IPCC AR4) Post et al. (2012) used the IPCC A1B global warming scenario output to transform historic daily climate records to provide future daily climate projections of P and PET that can used in a rainfall-runoff model. Compared to the global average temperature in 1990, the IPCC A1B scenario indicates a global temperature that is 1 °C higher in 2030 and 2 °C higher in 2070. This scenario is based upon: (i) very rapid economic growth, (ii) with global populations peaking mid-century and declining thereafter, and (iii) the rapid introduction of new and more efficient technologies with a balance across all energy sources (IPCC, 2007). Full details of the transformation of historic daily climate records using IPCC AR4 output are reported in Chiew et al. (2009) and Li et al. (2009).
Post et al. (2012) assessed the changes in P for the 15 GCMs and reported changes for large catchments such as the Manning River and Karuah River catchments, which form the broader context that the Gloucester subregion sits in (see Figure 7). Table 3 shows that for both catchments about two-thirds of the GCMs selected suggest there will be some decline in P. Taking account for the range of projections that may occur for the one-degree rise in temperatures (associated with 2030) there is approximately a –8%, –2% and 3% change in P projected for the dry extreme, median and wet extreme, respectively. For a 2-degree rise in temperatures (associated with 2070), these values are approximately –16%, –4% and 6%, respectively (Table 3).
Table 3 Summary of projected impacts of climate change on rainfall for the broad vicinity of the Gloucester subregion
Source data: Post et al. (2012; their Table 2)
To model future projects of runoff (Q), Post et al. (2012) used the future projections of daily P, along with a form of PET (specifically Morton’s wet environment areal formulation) as input to a lumped conceptual rainfall-runoff model called SIMHYD which utilises the Muskingum routing method (Chiew et al., 2009). Table 4 shows that the Post et al. (2012) modelling results suggest for a one-degree rise in temperatures (associated with 2030) there is approximately a –20%, –8% and 4% change in Q projected for the dry extreme, median and wet extreme, respectively. For a two-degree rise in temperatures (associated with 2070), these values are approximately –38%, –14% and 8%, respectively (Table 4). As noted previously the Gloucester basin is ‘equitant’ and so estimates of both P and PET are important for future projections of Q (McVicar et al., 2012b). Given this, use of Morton’s wet environment areal formulation of PET, which does not include wind speed in its formulation, means that the impact of declining rates of observed wind speed which are offsetting increasing air temperature enhancement of PET (Donohue et al., 2010; McVicar et al., 2012a; McVicar et al., 2012b) are not included in the resultant Q calculations. Hence the values presented in Table 4 are approximate projections only, as recent key process understanding is not encapsulated in the modelling.
Table 4 Summary of projected impacts of climate change on runoff for the broad vicinity of the Gloucester subregion
Source data: Post et al. (2012; their Table 3)
