| Summary: | In this paper, we investigate the impact of historical land cover change on climate extremes in eastern Australia by analysing data from an ensemble of model simulations using CSIRO AGCM. The model simulations were performed for two sets of prescribed land surface parameters representative of pre-European and modern-day land cover conditions. To evaluate the impact of historical land cover change on Australian regional climate, the CSIRO AGCM was used to complete two sets of model simulations (ensemble of 10 each) for the period 1951-2003. In this study, we used the CSIRO climate model consisting of atmospheric and land surface components forced by observed sea surface temperature and sea ice data for the period 1951-2003 (Rayner et al., 1996). This experimental set-up followed the design of the Climate of the 20th Century project (Folland et al., 2002) and allows for direct comparison between observed and model simulated ENSO events which are known to strongly influence Australian climate. The only difference between the experiments was the land surface characteristics for Australian continent used by the CSIRO model. The first set of model simulations used the modern-day and the second used the pre-European land cover characteristics. Outside Australia, the land cover characteristics were set at modern day conditions for both experiments. The modern-day land surface conditions were derived using data from the AVHRR satellite imagery for the period 1981 to 2001 at an 8 km spatial footprint (Lawrence, 2004). The monthly long-term average values of vegetation cover class, leaf area index, vegetation fraction and surface albedo were used as an input to the Simple Biosphere Model (SiB) derivation methods described in Sellers et al. (1986) to compute land surface characteristics used by the CSIRO climate model. Pre-clearing land surface parameters of vegetation fraction, leaf area index, surface albedo and stomatal resistance were generated by extrapolating the modern-day monthly values of remnant native vegetation to the pre-European coverage (see Lawrence, 2004). The extrapolation was performed for the Australian continent at an ~8×8 km resolution and aggregated to ~200×200 km resolution used by CSIRO AGCM using the approach of Shuttleworth, (1991), thereby ensuring the seasonal dynamics captured by satellite imagery were represented in pre-European parameters. The impact of land cover change on mean climate in Australia was described in McAlpine et al. (2007). The results showed a statistically significant increase in mean annual surface temperature and decrease in mean annual rainfall in southeast Australia. On a seasonal basis, the impact of land cover change was strongest during the summer season and was especially pronounced during strong El Niño events such as the 2002/03 event. In this paper, we focus on the impact of land cover change on the climate extremes by analysing the daily statistics of rainfall and temperature change over the period 1951-2003. To quantify the changes in annual distribution of daily rainfall and temperature, we computed the probability distribution functions (pdfs) of daily maximum surface temperature (tmax) and daily rainfall for selected locations in eastern Australia. In addition, the daily rainfall and temperature data was used to derive climate extreme indices of dry days (number of days with rainfall < 1 mm), daily rainfall intensity (total annual rainfall / number of rain days), rain days (number of days with rainfall 1 mm) and hot days (number of days with tmax ≥ 35ºC) (Frich et al., 2002). The analysis results showed statistically significant changes in the annual pdfs of rainfall and temperature in southeast Australia, which corresponds well with areas with largest fragmentation of pre-European vegetation cover. The fragmentation of vegetation resulted in an increase in the number of hot days, a decrease in daily rainfall intensity and a decrease in cumulative rainfall on rainy days in southeast Australia. These changes were especially pronounced during strong El Niño events.
|
|---|