Floods in Eastern Subtropical Argentina …
39
percentile interval [75th, 95th) of the local daily precipitation distribution of wet days
(RR>01.0 mm) in the period 1961–1990, following Donat et al. (2013). Accordingly,
R95pTOT was defined as annual total precipitation when RR falls into the percentile
interval [95th, 100th). Additionally, to complement annual mean field analysis,
monthly gridded precipitation data were used, provided by the Global Precipitation
Climatology Centre (GPCP) version 7, on a high 0.5° latitude–longitude resolution
grid (Schneider et al. 2011). To assess future climate changes in precipitation and its
extremes, under scenarios of anthropogenic climate change due to increasing GHGs,
historical and future projections generated from general climate models (GCMs)
were analyzed. The simulations corresponded to the Climate Model Intercomparison
Project phase 5 (CMIP5. Taylor et al. 2012). We computed multi-model ensemble
means for annual total precipitation (ATOTP) and R95pTOT extreme precipitation
using 8.5 W m-2 radiative forcing Representative Concentration Pathway (RCP8.5)
simulations from the historical (1961–1990) and future climate (2075–2099) experiments. Ensemble means correspond to outputs from four models, selected out of
fifteen models which provide daily data (Table A2). The four models, after validation
and adjustment for daily frequency distribution bias, achieve to properly represent
the historical climate within ESA, as shown by the Argentine report of the Third
National Communication to the United Nations Framework Convention on Climate
Change (TNC 2014). Simulated daily data are provided on a common 0.5º latitude
and longitude resolution grid (retrieved from http://3cn.cima.fcen.uba.ar/).
Long-term changes in monthly river discharges for the Parana at Corrientes and
Túnel Subfluvial gauge stations (see Fig. 1), provided by the National Hydrology
Agency (retrieved from http://bdhi.hidricosargentina.gov.ar/), were also examined.
Monthly gridded data of the terrestrial water storage anomalies (TWSA), as measured
remotely by NASA’s Gravity Recovery and Climate Experiment (GRACE) mission
were examined in the period 2003–2015 (Landerer and Swenson 2012) to complement the hydrological balance in the region. Likewise, information of forest cover
and its intra-decadal changes were assessed for Argentina in the period 1990–2015,
provided by the UN Statistical Division, Environmental Indicators.
Trends were estimated by means of the linear least square technique (Wilks 2011).
The relationship among variables was estimated using Pearson’s one-moment correlation (r, and associated explained variance, r
2 ), and its statistical significance was
computed using a Student’s t-distribution for a z-transformation of the correlation
(Wilks 2011).
Structural Flood Risks
In Argentina, over the past 50 years, 75 major flood events have been reported,
affecting around 13 million people and taking more than 500 lives. With the equivalent of USD 22.5bn lost since 1980 (USD 43.5bn after adjusting for the country’s GDP
growth), floods are the costliest natural catastrophe, representing 58% of economic
losses generated by natural catastrophes (Swiss Re 2016).
39
percentile interval [75th, 95th) of the local daily precipitation distribution of wet days
(RR>01.0 mm) in the period 1961–1990, following Donat et al. (2013). Accordingly,
R95pTOT was defined as annual total precipitation when RR falls into the percentile
interval [95th, 100th). Additionally, to complement annual mean field analysis,
monthly gridded precipitation data were used, provided by the Global Precipitation
Climatology Centre (GPCP) version 7, on a high 0.5° latitude–longitude resolution
grid (Schneider et al. 2011). To assess future climate changes in precipitation and its
extremes, under scenarios of anthropogenic climate change due to increasing GHGs,
historical and future projections generated from general climate models (GCMs)
were analyzed. The simulations corresponded to the Climate Model Intercomparison
Project phase 5 (CMIP5. Taylor et al. 2012). We computed multi-model ensemble
means for annual total precipitation (ATOTP) and R95pTOT extreme precipitation
using 8.5 W m-2 radiative forcing Representative Concentration Pathway (RCP8.5)
simulations from the historical (1961–1990) and future climate (2075–2099) experiments. Ensemble means correspond to outputs from four models, selected out of
fifteen models which provide daily data (Table A2). The four models, after validation
and adjustment for daily frequency distribution bias, achieve to properly represent
the historical climate within ESA, as shown by the Argentine report of the Third
National Communication to the United Nations Framework Convention on Climate
Change (TNC 2014). Simulated daily data are provided on a common 0.5º latitude
and longitude resolution grid (retrieved from http://3cn.cima.fcen.uba.ar/).
Long-term changes in monthly river discharges for the Parana at Corrientes and
Túnel Subfluvial gauge stations (see Fig. 1), provided by the National Hydrology
Agency (retrieved from http://bdhi.hidricosargentina.gov.ar/), were also examined.
Monthly gridded data of the terrestrial water storage anomalies (TWSA), as measured
remotely by NASA’s Gravity Recovery and Climate Experiment (GRACE) mission
were examined in the period 2003–2015 (Landerer and Swenson 2012) to complement the hydrological balance in the region. Likewise, information of forest cover
and its intra-decadal changes were assessed for Argentina in the period 1990–2015,
provided by the UN Statistical Division, Environmental Indicators.
Trends were estimated by means of the linear least square technique (Wilks 2011).
The relationship among variables was estimated using Pearson’s one-moment correlation (r, and associated explained variance, r
2 ), and its statistical significance was
computed using a Student’s t-distribution for a z-transformation of the correlation
(Wilks 2011).
Structural Flood Risks
In Argentina, over the past 50 years, 75 major flood events have been reported,
affecting around 13 million people and taking more than 500 lives. With the equivalent of USD 22.5bn lost since 1980 (USD 43.5bn after adjusting for the country’s GDP
growth), floods are the costliest natural catastrophe, representing 58% of economic
losses generated by natural catastrophes (Swiss Re 2016).