1. Introduction
Rainfall shows considerable temporal and spatial variability around the world (Easterling et al. 2000). Spatial variability in rainfall can provide vital information and insight that would be useful for adopting region-specific water management strategies. The global precipitation pattern indicates that the wet areas are becoming wetter and dry areas are becoming drier (Dore 2005). In the United States, the Northwest and Southwest show decreasing trends on precipitation amount, whereas the Midwest, Northeast, and parts of the South show increasing trends (Prein et al. 2016). A significant decreasing trend has been reported in the annual rainfall of the Mediterranean Sea basin, and a significant increasing trend in annual rainfall in central and northern Europe has been found (Caloiero et al. 2018). In the Arab regions, a general warming trend has been observed since the 1970s. The western part of the region exhibits a wetting trend, and the eastern part shows a drying trend (Donat et al. 2014). Wang et al. (2015) studied the spatial and temporal variations of annual precipitation in China during the period 1960–2010 and showed that there have been significant changes in rainfall in east, northwest, northeast, and southwest China and that the decadal changes of precipitation in east China were closely correlated with the East Asian summer monsoon and the atmospheric circulation. Saha et al. (2018) suggests that the trend of rainfall over different regions in India shows a dissimilar pattern. The trend of rainfall in different seasons in north and central India is closely related to the overall decreasing pattern of the country; peninsular India shows slight similarity, while northeast India shows a pattern different from the overall decreasing pattern in rainfall.
In the Southern Hemisphere, a study conducted in New Zealand by Caloiero (2014), looked at the spatial and temporal patterns of daily precipitation concentration using the precipitation concentration index (CI). The precipitation CI is used to examine the temporal heterogeneity of precipitation distribution by correlating the magnitude of precipitation events with the time of occurrence (Zhang et al. 2019). The results showed the different behavior of CI between the North Island and the South Island. The North Island and the eastern side of the South Island showed higher CI values, indicating critical rainfall concentration, and the western side of the South Island showed lower CI values. In Australia, the average rainfall has increased in the northwest and southeast (Collins and Della-Marta 1999; Suppiah and Hennessy 1998), whereas there is a large significant decline of rainfall in the southwest of Western Australia (SWWA) (Wright 1974; Yu and Neil 1993; Chowdhury et al. 2015).
The decrease of rainfall in SWWA has been well studied and documented in previous studies. They have examined the change in rainfall in relation to the change in rainfall depth or frequency (Pittock 1983; Yu and Neil 1993; Hennessy 1999; Smith 2004; Gallant et al. 2007). Rainfall volume as distinct from rainfall depth is crucial because the precipitation depth over an area such as a river basin helps to determine the water resources availability for that area. Changes in the rainfall volume have considerable implications for regional water resources planning and management (Arai et al. 2012). Because the rainfall volume is the product of rainfall depth and the wet area, the change in rainfall volume is the result of change in rainfall depth or that in wet area or both. Rainfall volume has significant applications in terrestrial hydrology and in global circulation models (Kebe et al. 2005). Previous studies that considered rainfall volume were concerned with the rainfall volume produced by tropical storms and their effects (Nogueira and Keim 2010) and the relationship between rainfall volume and vegetation (Lázaro et al. 2001). To our knowledge, no research has been undertaken to examine the change in rainfall volume and the wet area in the context of a regional decline in rainfall in SWWA.
The southern annular mode (SAM) is the north–south movement of the westerly wind belt that dominates the middle to higher latitudes of the Southern Hemisphere (Thompson and Wallace, 2000). Previous studies suggest that a positive SAM in a neutral El Niño–Southern Oscillation (ENSO) phase is the cause of the decrease in rainfall along the coast of SWWA (Raut et al. 2014). The inverse relationship between SAM and SWWA winter rainfall has also been established in previous studies (Ansell et al. 2000; Li et al. 2005). However, Feng et al. (2010) found that the positive and negative phases of SAM have negligible impact on SWWA winter rainfall if the year 1964, which was an extremely wet year, is excluded. Most rainfall trend studies, as in this study, are usually based on long-term rainfall data that can overlook the recent climate variations of a region. As a 30-yr period is the minimum duration required to define regional climatology, the period from 1989 to 2018 was analyzed to examine recent climate variations in SWWA.
The aim of this study is to obtain a better understanding of the changes in rainfall volume in SWWA during the period from 1911 to 2018. The objectives of this study are to 1) evaluate the rate of change in rainfall volume, wet area, and the average daily rainfall depth for different rainfall thresholds; 2) attribute changes in rainfall volume to that in the wet area and average daily rainfall depth; 3) compare the observed trend in rainfall volume and wet area with the expected trend; 4) map the change in areas having different number of rain days for different rainfall thresholds; 5) analyze and interpret the recent trend (1989–2018) in rainfall and trend variations in relation to the SAM.
2. Material and methods
The study area (265 952 km2) is the southwest of Western Australia, which lies south of 30°S and west of 120°E. This area receives a mean annual rainfall (MAR) between 200 and 1258 mm (Fig. 1) and has a Mediterranean climate with more than 50% of the rainfall in winter (June–August). The daily gridded rainfall data from the Bureau of Meteorology from 1911–2018 (108 years) were used for this study (http://www.bom.gov.au/climate/data-services/maps.shtml). The gridded rainfall data have been developed under the Australian Water Availability Project (AWAP) where the grids are computer generated from station rainfall data using a sophisticated analysis technique as detailed by Jones et al. (2009). Each grid cell covers an area of 0.05° latitude by 0.05° longitude. On the basis of the mean annual rainfall in SWWA, the area was classified into four zones. Areas with the mean annual rainfall between 200 and 400 mm are zone 1, between 400 and 600 mm are zone 2, between 600 and 1000 mm are zone 3, and greater than 1000 mm are zone 4 (Fig. 1). Table 1 gives the details of each zone. Note that these zones lie at varying distances from the coast, with zone 4 being closest to the coast and zone 1 being in general farthest inland. For each zone, the changes in rainfall volume, wet area, and average daily rainfall depth were determined using linear regression. The statistical significance of the trend was determined using the p value. A p value of less than 5% was considered to be statistically significant.
MAR in the southwest of Western Australia classified into four zones. The data were obtained online (http://www.bom.gov.au) for the period 1961–90.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
MAR in the southwest of Western Australia classified into four zones. The data were obtained online (http://www.bom.gov.au) for the period 1961–90.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
MAR in the southwest of Western Australia classified into four zones. The data were obtained online (http://www.bom.gov.au) for the period 1961–90.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
The area and MAR volume for the whole study area and the four zones for the period 1911–2018.
A day with rainfall that was greater than or equal to 0.2 mm was considered to be a wet day, a day with rainfall of ≥10 mm was considered to be a heavy precipitation day, and a day with rainfall of ≥30 mm was considered to be a very heavy precipitation day according to the Australian Bureau of Meteorology (http://www.bom.gov.au/climate/change/about/extremes.shtml). In this study we define and investigate the following categories: daily rainfall ≥ 0.2 mm day−1 was considered to be total rainfall, daily rainfall ≥ 10 mm day−1 was considered to be heavy rainfall, and daily rainfall ≥ 30 mm day−1 was considered to be very heavy rainfall. The rainfall volume, wet area, and average daily rainfall depth for each year were computed as follows:
The rainfall volume on a given day V was the sum of rainfall depths for all of the grid cells with rainfall ≥ r (where r = 0.2, 10, and 30 mm day−1) times average area of the grid cell (30.9103 km2). The annual rainfall volume was calculated as the sum of rainfall volumes for all days (rainfall ≥ r mm) in the year, given in teraliters (TL).
The wet area on a given day was the total number of cells with rainfall ≥ r times the average area of the grid cell. The average wet area for a year A was calculated as the average of all daily wet areas in a year, given in kilometers squared.
The average daily rainfall depth D was the sum of rainfall depths for all grid cells with rainfall ≥ r in the year divided by the total number of cells with rainfall ≥ r in that year, given in millimeters.
Let V be the annual volume of rainfall of all grid cells with rainfall ≥ 0.2 mm day−1 and let Vr be the annual volume of rainfall of all grid cells with rainfall ≥ r (where r = 10 and 30 mm day−1). For each threshold r, let η be the ratio of the mean Vr to the mean V. As r increases, the contribution of Vr to total V would decrease or so would η. If V and Vr are perfectly correlated and proportional, that is, Vr(i) = ηV(i) for each year i, one would expect that the rate of change in Vr βr, would be exactly ηβ, where β is the rate of change in V. Thus, we have a null hypothesis that βr = ηβ to test whether rainfall above a threshold varies in a similar fashion in comparison with that in total rainfall volume. In other words, we could compare the observed rate of change in Vr, that is, βr, with the expected rate of change ηβ for different rainfall thresholds. Similarly, the observed rate of change in wet area was compared with the expected rate of change in wet area for different daily rainfall thresholds.
The number of days with rainfall ≥ 0.2 mm and rainfall ≥ 10 mm in a year for each grid cell in the whole study area was determined and shown as maps for each decade. These maps help identify regions where the change in the number of rain days was most pronounced. The four zones based on the mean annual rainfall were not considered in the study of rate of change in number of rain days. Instead, new regions were identified throughout the study area based on the number of rain days and the rate of change in the area of these regions were calculated.
To determine the recent changes in rainfall and their relation to the SAM, the total annual rainfall was calculated for the 0.2, 10, and 30 mm day−1 thresholds for the periods 1957–2018, 1957–88, and 1989–2018, respectively. The trend in the SAM and rainfall for these periods were calculated for the whole region and the four zones. Pearson correlation was used to find the statistical relationships between SAM and rainfall in the study area. The SAM index was extracted for the period 1957–2018 from Marshall (2003).
3. Results
a. Trend in rainfall volume, wet area, and average daily rainfall depth
The rainfall volume, wet area, and average daily rainfall depth were calculated for the whole area and the four zones separately for total rainfall (rainfall ≥ 0.2 mm day−1), heavy rainfall (rainfall ≥ 10 mm day−1), and very heavy rainfall (rainfall ≥ 30 mm day−1), respectively, for the period 1911–2018. Figures 2a and 2b show an example of the time series of annual rainfall volume and the average wet area for zone 4 (MAR ≥ 1000 mm) for the heavy rainfall (rainfall ≥ 10 mm day−1). A decreasing pattern was observed in rainfall volume and wet area in zone 4. Figure 2c represents the wet area (km2) for heavy rainfall (rainfall ≥ 10 mm day−1) in zone 4 for each day of the year 1957. The average of these daily values (1586.25 km2 in 1957) was taken as the average wet area for the particular year.
Time series of (a) annual rainfall volume, (b) annual wet area, and (c) daily wet area for 1957 in zone 4 for daily rainfall intensity > 10 mm as an example. Dotted lines represent a decreasing trend of 0.23 TL (10 yr)−1 for rainfall volume and 31.04 km2 (10 yr)−1 for wet area.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Time series of (a) annual rainfall volume, (b) annual wet area, and (c) daily wet area for 1957 in zone 4 for daily rainfall intensity > 10 mm as an example. Dotted lines represent a decreasing trend of 0.23 TL (10 yr)−1 for rainfall volume and 31.04 km2 (10 yr)−1 for wet area.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Time series of (a) annual rainfall volume, (b) annual wet area, and (c) daily wet area for 1957 in zone 4 for daily rainfall intensity > 10 mm as an example. Dotted lines represent a decreasing trend of 0.23 TL (10 yr)−1 for rainfall volume and 31.04 km2 (10 yr)−1 for wet area.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
1) Trend in rainfall volume
Figure 3 presents the results of trend analysis of the rainfall volume, for the whole study area and the four zones separately. The whole region showed a significant decreasing trend for the total rainfall (rainfall ≥ 0.2 mm day−1). Zone 1, with a MAR of 200–400 mm and which lies farther inland, showed an increasing trend in rainfall volume for total, heavy, and very heavy rainfall (rainfall ≥ 0.2, 10, and 30 mm day−1), although the trend was not statistically significant for all thresholds. Zone 2 (400 ≤ MAR ≤ 600 mm) showed a significant decreasing trend in rainfall volume for total rainfall (rainfall ≥ 0.2 mm day−1). Zone 3 (600 ≤ MAR ≤ 1000 mm) showed a significant decreasing trend in rainfall volume for total (rainfall ≥ 0.2 mm day−1) and heavy rainfall (rainfall ≥ 10 mm day−1), and zone 4 showed a significant decreasing trend for all thresholds. In general, the decreasing trend in rainfall volume was predominant in the coastal areas (zone 3 and 4), whereas the inland area (zone 1) showed an increasing trend, although this trend was not statistically significant.
The trend in the rainfall volume for the whole study area and four zones with different daily rainfall thresholds. The asterisk indicates a trend at 5% significance.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
The trend in the rainfall volume for the whole study area and four zones with different daily rainfall thresholds. The asterisk indicates a trend at 5% significance.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
The trend in the rainfall volume for the whole study area and four zones with different daily rainfall thresholds. The asterisk indicates a trend at 5% significance.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
2) Trend in wet area
Figure 4 presents the results of trend analysis of the average wet area for the region and for individual zones. Significant decreasing trend in wet area was observed for total rainfall (rainfall ≥ 0.2 mm day−1) for the whole region and in all zones. Zone 1 showed an increasing trend in wet area for the heavy (rainfall ≥ 10 mm day−1) and very heavy rainfall (rainfall ≥ 30 mm day−1) although this trend was not significant. Zone 3 and 4 showed significant decrease in wet area with heavy rainfall (rainfall ≥ 10 mm day−1), and zone 4 had a significant decrease in wet area with very heavy rainfall than (rainfall ≥ 30 mm day−1). These results suggested that the wet area has been shrinking in size for the study period (1911–2018) in the regions of high MAR (≥ 600 mm). This could be one of the reasons for the observed decreasing trend in rainfall in the study area. A net decrease in the wet area for the whole region was observed, even though there was an increase in the wet area farther inland, that is, zone 1, because this increase was more than offset by the concurrent decrease in other zones.
As in Fig. 3, but for the trend in the wet area.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
As in Fig. 3, but for the trend in the wet area.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
As in Fig. 3, but for the trend in the wet area.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
3) Trend in average daily rainfall depth
Figure 5 presents the trend in the average daily rainfall depth in the study area. The mean rainfall depth increases as the rainfall threshold increases. Significant trends were observed only for the total rainfall (rainfall ≥ 0.2 mm day−1) for the whole region and in zone 1. All three thresholds showed an insignificant increase in average daily rainfall depth for the whole region and all zones except for heavy rainfall (rainfall ≥ 10 mm day−1) in zone 4 and very heavy rainfall (rainfall ≥ 30 mm day−1) in zone 1.
As in Fig. 3, but for the trend in the rainfall depth.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
As in Fig. 3, but for the trend in the rainfall depth.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
As in Fig. 3, but for the trend in the rainfall depth.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
To summarize, the total rainfall (rainfall ≥ 0.2 mm day−1) showed a significant decreasing trend in rainfall volume and wet area, and an increasing trend in the average daily rainfall depth for the whole region and in most zones. For the heavy rainfall (rainfall ≥ 10 mm day−1), a significant decreasing trend in rainfall volume and wet area was observed in the coastal regions (zone 3 and 4) while an increasing trend was observed in the inland areas (zone 1). The average daily rainfall depth for heavy rainfall showed an increasing trend in all zones except zone 4. The very heavy rainfall (rainfall ≥ 30 mm day−1) in the coastal region with MAR ≥ 1000 mm (zone 4) showed a significant decrease in rainfall volume and wet area and a nonsignificant increasing trend in the average daily rainfall depth in all zones except zone 1.
b. Contribution of change in wet area and average daily rainfall depth to the change in rainfall volume
The rate of change in wet area contributed most to the rate of change in rainfall volume, while the rate of change in rainfall depth only played a minor role. Table 2 gives the range of percentage contribution of rate of change in wet area and average daily rainfall depth to the rate of change in rainfall volume. For the whole region, for all thresholds, more than 55% of the rate of change in volume was caused by a rate of change in wet area. In zone 1, for the heavy (rainfall ≥ 10 mm day−1) and very heavy rainfall (rainfall ≥ 30 mm day−1), more than 77% of the rate of change in volume was due to the rate of change in wet area, while for total rainfall (rainfall ≥ 0.2 mm day−1), 54% of the rate of change in volume was due to rate of change in average daily rainfall depth. In zone 2, more than 50% of the rate of change in rainfall volume and in the coastal region (zones 3 and 4), more than 84% of the rate of change in volume, was due to the rate of change in wet area. In general, it was observed that for most regions the rate of change in volume was caused mainly by the rate of change in wet area rather than the rate of change in average daily rainfall depth.
Percentage contribution of the rate of change in the wet area and that in the average daily rainfall depth to the rate of change in rainfall volume.
c. Observed versus expected trends for rainfall volume and wet area
The observed trend was compared with the expected trend in rainfall volume and wet area as described in section 2. Figures 6a and 6b show the observed and expected trends of rainfall volume for the heavy (rainfall ≥ 10 mm day−1) and very heavy rainfall (rainfall ≥ 30 mm day−1), respectively, for the whole region and the four zones considered for the study. The observed decreasing trend of rainfall volume for the heavy rainfall (rainfall ≥ 10 mm day−1), as given in Fig. 6a, was lower than the expected decreasing trend for the whole region and zones 2, 3, and 4, whereas in zone 1 the observed increasing trend was higher than the expected increasing trend. For the whole region and zone 2, a decreasing trend in the rainfall volume of very heavy rainfall (rainfall ≥ 30 mm day−1) was expected but an increasing trend was observed instead (Fig. 6b). In zone 1 the observed increasing trend was higher than the expected increasing trend, in zone 3 the observed decreasing trend was lower than the expected decreasing trend, and in zone 4 the observed decreasing trend was higher than the expected decreasing trend.
Trends in (a),(b) rainfall volume and (c),(d) wet area for the rainfall threshold of (left) 10 and (right) 30 mm day−1. In each plot, the dashed line indicates expected trend and the bars indicate observed trends.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Trends in (a),(b) rainfall volume and (c),(d) wet area for the rainfall threshold of (left) 10 and (right) 30 mm day−1. In each plot, the dashed line indicates expected trend and the bars indicate observed trends.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Trends in (a),(b) rainfall volume and (c),(d) wet area for the rainfall threshold of (left) 10 and (right) 30 mm day−1. In each plot, the dashed line indicates expected trend and the bars indicate observed trends.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Figures 6c and 6d show the observed and expected trends of wet area for the heavy (rainfall ≥ 10 mm day−1) and very heavy rainfall (rainfall ≥ 30 mm day−1) for the region and the four zones. The decreasing trend of wet area for the heavy rainfall (rainfall ≥ 10 mm day−1) was lower than the expected decreasing trend for the whole region and zones 2, 3, and 4 (Fig. 6c), whereas in zone 1 a decreasing trend was expected but an increasing trend was observed. For the very heavy rainfall (rainfall ≥ 30 mm day−1), the observed decreasing trend in the wet area for the whole region and zones 2 and 3 was lower than expected (Fig. 6d). In zone 1 a decreasing trend was expected but an increasing trend was observed, whereas in zone 4 the observed decreasing trend was higher than the expected decreasing trend.
In general, in the inland region (zone 1; 200 ≤ MAR ≤ 400 mm) the observed increasing trend in rainfall volume was higher than the expected increasing trend and for wet area a decreasing trend was expected but an increasing trend was observed. In the coastal region (zone 4; 1000 ≤ MAR ≤ 1258 mm), the observed decreasing trend in rainfall volume and wet area was higher than the expected decreasing trend.
d. Trend in number of rain days
Figure 7 represents the change in the wet area corresponding to the number of days with rainfall greater than or equal to 0.2 mm day−1.
Interdecadal spatial variation in rain days with rain ≥ 0.2 mm day−1.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Interdecadal spatial variation in rain days with rain ≥ 0.2 mm day−1.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Interdecadal spatial variation in rain days with rain ≥ 0.2 mm day−1.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
The southern and western regions of the study area received more rain days than the northern and eastern parts of the study area (Fig. 7). Table 3 compares the fraction of the total area, receiving different classes of rain days for the first and last decade. The region in the southwest coast receiving greater than or equal to 200 rain days yr−1, which covered 6.5% of the study area in 1910s, decreased to 0% by the 2010s (Table 3). The wet area with fewer than 120 rain days yr−1, in the northeast of the study area, increased by 28%, whereas the region in the southwest receiving greater than or equal to 120 rain days yr−1 decreased by 28%.
Comparison of the fractional area in terms of the number of rain days per year between the 1910s and 2010s.
Figure 8 gives the change in wet area corresponding to the number of days with rainfall greater than or equal to 10 mm day−1. Overall, the west and southwest of the study area received the highest number of days with rainfall greater than or equal to 10 mm day−1. This region in the west and southwest, with greater than or equal to 30 rain days yr−1, covered 12.8% of the total area in the 1910s but decreased to 6.5% in the 2010s. The area with fewer than 30 rain days yr−1 that lies inland increased in area by 6.5% by the 2010s (Table 3). The region in the west that received greater than or equal to 40 rain days yr−1 disappeared by the end of the study period, and the region in the south with greater than or equal to 40 rain days yr−1 gradually decreased in area.
As in Fig. 7, but for rain days with rain ≥ 10 mm day−1.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
As in Fig. 7, but for rain days with rain ≥ 10 mm day−1.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
As in Fig. 7, but for rain days with rain ≥ 10 mm day−1.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Table 4 gives the trend in wet area corresponding to the number of days with rainfall ≥ 0.2 mm day−1 and rainfall ≥ 10 mm day−1, respectively. The wet area with rainfall greater than or equal to 0.2 mm day−1 and with fewer than 120 rain days yr−1 showed a significant increasing trend while the region with more than 120 rain days yr−1 showed a significant decreasing trend. The wet area with rainfall greater than or equal to 10 mm day−1 and fewer than 10 days of rainfall showed a decreasing trend. The wet area with 10–20 rain days yr−1 showed an increasing trend while the wet area with more than 20 rain days yr−1 showed a decreasing trend. Mostly, the wet area with the higher number of rain days showed a significant decreasing trend in both of the conditions that were considered while the wet area with lower number of rain days showed an increasing trend.
Trend in the wet areas. An asterisk indicates significance at 5% (positive trends indicate that the wet area is expanding, and negative trends indicate that the wet area is contracting).
e. Recent changes in rainfall in relation to the SAM
Trend analysis for the last 30 years showed that in the coastal regions (zones 3 and 4), the decreasing trend has reduced in magnitude for the total annual rainfall (rainfall ≥0.2 mm day−1), while heavy rainfall (rainfall ≥10 mm day−1) and very heavy rainfall (rainfall ≥30 mm day−1) showed an increasing trend, with no concurrent change in SAM. The rate of increase in SAM over the past 30 years (1989–2018) was 0.54 (10 yr)−1, which was similar to the long-term rate of increase [0.55 (10 yr)−1] (Fig. 9a). In the coast (zones 3 and 4), the total annual rainfall decreased over the whole period (1957–2018) and showed a negative correlation (−0.29) with the SAM. This is consistent with the findings of Li et al. (2005) and Hope et al. (2010) who suggested an inverse relation between the SAM and rainfall. However, it was observed that the rate of decrease in rainfall has substantially reduced in magnitude over the past 30 years (1989–2018) relative to the previous 32 years (1957–88) (Fig. 9b). This was true for the coastal region (zones 3 and 4) for all rainfall thresholds, as presented in Table 5. For some, the trend has been reversed, that is, rainfall has in fact increased over the past 30 years. For example, in zone 4, during the period from 1957 to 1988, the heavy rainfall significantly decreased at the rate of 61.5 mm (10 yr)−1, whereas in the last 30 years the heavy rainfall showed an increasing trend [15.8 mm (10 yr)−1] as presented in Fig. 9c. This change in the way rainfall varied over the past 30 years suggests that the relationship between SAM and rainfall has changed over the recent years and that other mechanisms are needed to explain the recent rainfall trend and variations in coastal regions of SWWA.
Time series of (a) SAM, (b) total annual rainfall (0.2 mm day−1), and (c) annual heavy rainfall (10 mm day−1) for the period 1957–2018 in zone 4. In each plot, the dashed line represents the trend line.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Time series of (a) SAM, (b) total annual rainfall (0.2 mm day−1), and (c) annual heavy rainfall (10 mm day−1) for the period 1957–2018 in zone 4. In each plot, the dashed line represents the trend line.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Time series of (a) SAM, (b) total annual rainfall (0.2 mm day−1), and (c) annual heavy rainfall (10 mm day−1) for the period 1957–2018 in zone 4. In each plot, the dashed line represents the trend line.
Citation: Journal of Applied Meteorology and Climatology 60, 1; 10.1175/JAMC-D-20-0034.1
Comparison of trends in the total, heavy, and very heavy annual rainfall in zones 3 and 4 during the period 1957–88 and 1989–2018. An asterisk indicates significance at 5%.
4. Discussion
In SWWA, the region of high rainfall near the coast (zones 3 and 4) with a larger number of rain days was observed to become drier due to a significantly decreasing trend in the wet area. On the other hand, the region farther inland (zone 1) with low MAR and a lower number of rain days was observed to become wetter due to an increasing trend in the wet area, although this trend was not significant. SWWA receives most of its rainfall from the frontal systems (Charles et al. 2010; Pook et al. 2012), but since 1974, there have been fewer and weaker cold fronts (Frederiksen and Frederiksen 2007), which could explain the reduced wet area in SWWA. The decline in rainfall in the coastal regions has also been associated with the decrease in the frequency of strong westerly fronts (Nicholls et al. 1997; Hope et al. 2006). The observed decrease in wet area could also be attributed to the southward shift in synoptic circulations, associated with rain-bearing fronts, since the 1970s (Hope et al. 2006). The increasing trend in the inland region can be attributed to an increase in the frequency of easterly troughs in December, January, and February in the inland region of the study area (Berry et al. 2011). Further, the positive phase of SAM in a neutral ENSO phase could cause the region to experience reduced rainfall from the westerly fronts and increased rainfall from easterly troughs (Raut et al. 2014).
Previous studies have suggested the role of the SAM in the decreasing pattern of rainfall in SWWA (Li et al. 2005; Raut et al. 2014). However, in this study, it was observed that in the recent years (1989–2018), rate of decrease in rainfall has been reduced, and heavy rainfall in fact has increased in the coastal region of SWWA, while, there was no concurrent change in SAM index over the same period. This indicates that the relationship between SAM and rainfall has changed and other climate drivers could be the cause of the recent rainfall trend and variations in the coastal regions of SWWA.
5. Conclusions
This study aimed at determining the changes in rainfall volume, wet area, average daily rainfall depth, and the number of wet days in SWWA during the period 1911–2018. The study dealt with identifying the rainfall changes in four zones, classified based on the mean annual rainfall and located at varying distances away from the coast, with zone 4 being the closest to the coast with the highest mean annual rainfall, while zone 1 lay farther inland with the lowest mean annual rainfall. Also, in the study, rainfall at different intensities was considered to have a better understanding of the variation of heavy rainfall above certain threshold in relation to an overall decreasing trend in rainfall volume. Recent trend (1989–2018) in the total, heavy, and very heavy rainfall were analyzed, and this was explained relative to the SAM. The key conclusions are given below:
The significant decreasing trend in rainfall volume and wet area was predominant in coastal regions (zone 3 and 4), while the inland region (zone 1) showed an increasing trend although this trend was not statistically significant for all thresholds considered. In the inland region (zone 1), the observed increasing trend in rainfall volume was greater than the expected increasing trend. A decreasing trend in wet area was expected but an increasing trend was observed. In the coastal region (zone 4), observed decreasing trend in volume and wet area was greater than the expected decreasing trend.
All three thresholds showed an increasing trend in average daily rainfall depth in the whole region and all zones except for heavy rainfall (rainfall ≥ 10 mm day−1) in zone 4 and very heavy rainfall (rainfall ≥ 30 mm day−1) in zone 1 where a decreasing trend was observed.
More than 50% of the rate of change in rainfall volume in the whole region, zone 1 and 2 was attributed to the change in wet area except the for the total rainfall (rainfall ≥ 0.2 mm day−1) in zone 1 where 54% of the rate of change in rainfall volume was due to the rate of change in average daily rainfall depth. In the coastal regions (zone 3 and 4), more than 84% of the rate of change in rainfall volume was due to the rate of change in wet area.
The region near the coast with a higher number of rain days per year showed a decreasing trend while the inland region with a lower number of rain days showed an increasing trend for rainfall ≥ 0.2 mm day−1 and rainfall ≥ 10 mm day−1.
The decreasing trend in total rainfall has reduced in magnitude, and heavy rainfall in fact has increased over the past 30 years in the coastal region of SWWA, while there was no concurrent change in SAM index over the same period.
The study showed that the rainfall volume in coastal areas is showing a significant decreasing trend. This indicates the need for strategic planning in the management and allocation of the available water in a region of decreasing water availability. This study was done using gridded rainfall data that are interpolated from station data. In SWWA, the density of stations is higher toward the coast compared to the inland regions. This is a limitation of this work, as the data quality in the region with fewer stations would be lower than the coastal region. Similarly, the number of rain days reported in this study will probably be higher than the actual number of rain days as the number of rain days is sensitive to the density of observations and as it takes only one station to record rainfall for surrounding grids to have an interpolated rainfall amount (Charles et al. 2010). Future studies could look into the effects of the observed decrease in rainfall volume and wet area on the water supply and agricultural sectors. Also, the changes in the climate drivers that influence the rainfall of SWWA could be further investigated to explain the recent changes in rainfall in the coastal regions of SWWA.
Acknowledgments
The authors acknowledge Griffith University for the financial support provided through the Griffith University International Postgraduate Research Scholarship (GUIPRS) and the Griffith University Postgraduate Research Scholarship (GUPRS). The authors also thank Zhuo Cheng for her valuable support during this study. The authors declare no competing interests.
REFERENCES
Ansell, T. J., C. J. C. Reason, I. N. Smith, and K. Keay, 2000: Evidence for decadal variability in southern Australian rainfall and relationships with regional pressure and sea surface temperature. Int. J. Climatol., 20, 1113–1129, https://doi.org/10.1002/1097-0088(200008)20:10<1113::AID-JOC531>3.0.CO;2-N.
Arai, F. K., S. B. Pereira, and G. G. Gonçalves, 2012: Characterization of water availability in a hydrographic basin. Eng. Agric., 32, 591–601, https://doi.org/10.1590/S0100-69162012000300018.
Berry, G., C. Jakob, and M. Reeder, 2011: Recent global trends in atmospheric fronts. Geophys. Res. Lett., 38, L21812, https://doi.org/10.1029/2011GL049481.
Caloiero, T., 2014: Analysis of daily rainfall concentration in New Zealand. Nat. Hazards, 72, 389–404, https://doi.org/10.1007/s11069-013-1015-1.
Caloiero, T., P. Caloiero, and F. Frustaci, 2018: Long-term precipitation trend analysis in Europe and in the Mediterranean basin. Water Environ. J., 32, 433–445, https://doi.org/10.1111/wej.12346.
Charles, S., and Coauthors, 2010: Climate analyzes for South-West Western Australia: A report to the Australian Government from the CSIRO South-West Western Australia Sustainable Yields Project. CSIRO Rep., 83 pp.
Chowdhury, R. K., S. Beecham, J. Boland, and J. Piantadosi, 2015: Understanding South Australian rainfall trends and step changes. Int. J. Climatol., 35, 348–360, https://doi.org/10.1002/joc.3982.
Collins, D. A., and P. M. Della-Marta, 1999: Annual climate summary 1998: Australia’s warmest year on record. Aust. Meteor. Mag., 48, 273–283.
Donat, M. G., and Coauthors, 2014: Changes in extreme temperature and precipitation in the Arab region: Long-term trends and variability related to ENSO and NAO. Int. J. Climatol., 34, 581–592, https://doi.org/10.1002/joc.3707.
Dore, M. H., 2005: Climate change and changes in global precipitation patterns: What do we know? Environ. Int., 31, 1167–1181, https://doi.org/10.1016/j.envint.2005.03.004.
Easterling, D. R., G. A. Meehl, C. Parmesan, S. A. Changnon, T. R. Karl, and L. O. Mearns, 2000: Climate extremes: Observations, modeling, and impacts. Science, 289, 2068–2074, https://doi.org/10.1126/science.289.5487.2068.
Feng, J., J. Li, and Y. Li, 2010: Is there a relationship between the SAM and southwest Western Australian winter rainfall? J. Climate, 23, 6082–6089, https://doi.org/10.1175/2010JCLI3667.1.
Frederiksen, J. S., and C. S. Frederiksen, 2007: Interdecadal changes in Southern Hemisphere winter storm track modes. Tellus, 59A, 599–617, https://doi.org/10.1111/j.1600-0870.2007.00264.x.
Gallant, A. J., K. J. Hennessy, and J. Risbey, 2007: Trends in rainfall indices for six Australian regions: 1910-2005. Aust. Meteor. Mag., 56, 223–241.
Hennessy, K. J., 1999: Australian rainfall changes, 1910-1995. Aust. Meteor. Mag., 48, 1–13.
Hope, P. K., W. Drosdowsky, and N. Nicholls, 2006: Shifts in the synoptic systems influencing southwest Western Australia. Climate Dyn., 26, 751–764, https://doi.org/10.1007/s00382-006-0115-y.
Hope, P. K., B. Timbal, and R. Fawcett, 2010: Associations between rainfall variability in the southwest and southeast of Australia and their evolution through time. Int. J. Climatol., 30, 1360–1371, https://doi.org/10.1002/joc.1964.
Jones, D. A., W. Wang, and R. Fawcett, 2009: High-quality spatial climate data-sets for Australia. Aust. Meteor. Oceanogr. J., 58, 233–248, http://www.bom.gov.au/jshess/docs/2009/jones_hres.pdf.
Kebe, C. M. F., H. Sauvageot, and A. Nzeukou, 2005: The relation between rainfall and area–time integrals at the transition from an arid to an equatorial climate. J. Climate, 18, 3806–3819, https://doi.org/10.1175/JCLI3451.1.
Lázaro, R., F. S. Rodrigo, L. Gutiérrez, F. Domingo, and J. Puigdefábregas, 2001: Analysis of a 30-year rainfall record (1967–1997) in semi-arid SE Spain for implications on vegetation. J. Arid Environ., 48, 373–395, https://doi.org/10.1006/jare.2000.0755.
Li, Y., W. Cai, and E. P. Campbell, 2005: Statistical modeling of extreme rainfall in southwest Western Australia. J. Climate, 18, 852–863, https://doi.org/10.1175/JCLI-3296.1.
Marshall, G. J., 2003: Trends in the southern annular mode from observations and reanalyses. J. Climate, 16, 4134–4143, https://doi.org/10.1175/1520-0442(2003)016<4134:TITSAM>2.0.CO;2.
Nicholls, N., L. Chambers, M. Haylock, C. S. Frederiksen, D. Jones, and W. Drosdowsky, 1997: Climate variability and predictability for south-west Western Australia. IOCI Phase 1 Rep., 52 pp.
Nogueira, R. C., and B. D. Keim, 2010: Annual volume and area variations in tropical cyclone rainfall over the eastern United States. J. Climate, 23, 4363–4374, https://doi.org/10.1175/2010JCLI3443.1.
Pittock, A. B., 1983: Recent climatic change in Australia: Implications for a CO2-warmed earth. Climatic Change, 5, 321–340, https://doi.org/10.1007/BF02423529.
Pook, M. J., J. S. Risbey, and P. C. McIntosh, 2012: The synoptic climatology of cool-season rainfall in the central wheatbelt of Western Australia. Mon. Wea. Rev., 140, 28–43, https://doi.org/10.1175/MWR-D-11-00048.1.
Prein, A. F., G. J. Holland, R. M. Rasmussen, M. P. Clark, and M. R. Tye, 2016: Running dry: The US Southwest’s drift into a drier climate state. Geophys. Res. Lett., 43, 1272–1279, https://doi.org/10.1002/2015GL066727.
Raut, B. A., C. Jakob, and M. J. Reeder, 2014: Rainfall changes over southwestern Australia and their relationship to the southern annular mode and ENSO. J. Climate, 27, 5801–5814, https://doi.org/10.1175/JCLI-D-13-00773.1.
Saha, S., D. Chakraborty, R. K. Paul, S. Samanta, and S. B. Singh, 2018: Disparity in rainfall trend and patterns among different regions: Analysis of 158 years’ time series of rainfall dataset across India. Theor. Appl. Climatol., 134, 381–395, https://doi.org/10.1007/s00704-017-2280-9.
Smith, I., 2004: An assessment of recent trends in Australian rainfall. Aust. Meteor. Mag., 53, 163–173.
Suppiah, R., and K. J. Hennessy, 1998: Trends in total rainfall, heavy rain events and number of dry days in Australia, 1910–1990. Int. J. Climatol., 18, 1141–1164, https://doi.org/10.1002/(SICI)1097-0088(199808)18:10<1141::AID-JOC286>3.0.CO;2-P.
Thompson, D. W., and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13, 1000–1016, https://doi.org/10.1175/1520-0442(2000)013<1000:AMITEC>2.0.CO;2.
Wang, Y., Y. Ding, B. Ye, F. Liu, and J. Wang, 2013: Contributions of climate and human activities to changes in runoff of the Yellow and Yangtze Rivers from 1950 to 2008. Sci. China Earth Sci., 56, 1398–1412, https://doi.org/10.1007/s11430-012-4505-1.
Wang, Y., X. Chen, and F. Yan, 2015: Spatial and temporal variations of annual precipitation during 1960–2010 in China. Quat. Int., 380–381, 5–13, https://doi.org/10.1016/j.quaint.2014.12.047.
Wright, P. B., 1974: Seasonal rainfall in southwestern Australia and the general circulation. Mon. Wea. Rev., 102, 219–232, https://doi.org/10.1175/1520-0493(1974)102<0219:SRISAA>2.0.CO;2.
Yu, B., and D. T. Neil, 1993: Long-term variations in regional rainfall in the south-west of Western Australia and the difference between average and high intensity rainfalls. Int. J. Climatol., 13, 77–88, https://doi.org/10.1002/joc.3370130106.
Zhang, K., Y. Yao, X. Qian, and J. Wang, 2019: Various characteristics of precipitation concentration index and its cause analysis in China between 1960 and 2016. Int. J. Climatol., 39, 4648–4658, https://doi.org/10.1002/joc.6092.
