Introduction:

From different studies and witnessed abnormalities around the globe, it is now clear that climate change has brought and will bring vulnerabilities. CO2 in the atmosphere has set the record in 2018 since preindustrial era (1850-1900) (Poloczanska, Mintenbeck et al. 2018). Some individual sites indicate that the level has been increasing for the year of 2019. Consequently, global mean temperature has been rising 1.5℃ above preindustrial (1850-1900) era (Poloczanska, Mintenbeck et al. 2018). Other key indicators of the critical situation are Sea Level Rise (SLR) and Sea Ice Extent (SIC); both are the direct consequence of greenhouse gas (GHG) increase in the atmosphere. SLR has hit the record in 2019 of 3.2 mm/year during the 1993-2019 period and also Arctic extent has been decreasing (Poloczanska, Mintenbeck et al. 2018). On the local scale, the warming has consistent with the weather and climate variabilities related to climate change; North America has been unusually cold due to the crisis (Poloczanska, Mintenbeck et al. 2018). These changes have changed and will continue to change the climate, making the weather harsh. Extreme precipitations, more frequent hurricanes, intense tropical cyclones, unexpected thunderstorms and tornados, sever cold breaks, prolonged droughts, and seasonal timing shifts are expected to be common (Groisman, Knight et al. 2005; Diffenbaugh, Scherer et al. 2013; Emanuel 2013; Diffenbaugh, Swain et al. 2015; Gao, Leung et al. 2015; Sobel, Camargo et al. 2016). These threats not only target human communities but also do threaten ecosystems functionality.
Climate models have proven to be reliable enough to take actions against climate change reverse impacts. A projected hydrological model has the ability to plan for management practices in a given watershed. Hydrological projections have been studied by researchers around the globe at local and global scales. Significant changes in hydrological regimes in most part of the land surface of the planet will be likely to occur by midcentury (2050) or they are sensitive to global mean temperature below 2℃ (Arnell and Gosling 2013). AR4 (Field, Barros et al. 2012) of IPCC reported low confidence on the anthropogenic climate change footprints in flood records, which was due to limited instrumental records in terms of space and time. However, regarding impact on hydrological cycle like precipitation and snowmelt, and earlier spring peak flows, the report showed medium to high confidence which the latter was very likely (Field, Barros et al. 2012; Georgakakos, Fleming et al. 2014). Hydrological projections indicate sensitivity of flood frequency and climate change will be the partial reason for increasing flood likelihood on the global scale (Cisneros, BE et al. 2014). Climate change will adversely change streamflow and water quality and consequently will jeopardize freshwater ecosystems (Cisneros, BE et al. 2014). The projections also reveal medium to high level of confidence in posing risk to potable water (Cisneros, BE et al. 2014).
In Southeast of the US, seasonal drying has been observed for spring, fall , and winter and in summer the soil moisture has increase during 1988-2010 (Georgakakos, Fleming et al. 2014). Although potential evapotranspiration (PET) has projected to increase, evapotranspiration (ET) as the 2nd largest component of the hydrological cycle requires further studies to see possible implication of the causes like soil moisture in future ET trends (Georgakakos, Fleming et al. 2014). Projections for annual runoff and consequent stream flow in the Southeast indicate declines, which is consistent with long-term (multi-seasonal) droughts that are projected for the Southeast (Georgakakos, Fleming et al. 2014).
River floods are more complex to be considered as the direct result of the heavy precipitation and topography, soil moisture, channel condition, and anthropogenic influences are thought to play the key roles (Georgakakos, Fleming et al. 2014). River floods have been decreasing in most part of the Southeast at least 6% percent per decade (Villarini, Serinaldi et al. 2009; Georgakakos, Fleming et al. 2014). Areas close to the Golf Coast have been hit by hurricanes several times in recent years. Catchment specific characteristics like seasonality and storm frequencies have implications in the flood peaks (Villarini and Smith 2010). Since it is difficult to carry out a reliable projection for very heavy rainfall, it is necessary to study hydrological feedback of the area to the storms. Few studies like Ge Sun et.al., (2013) have studied the impact of climate change on the entire southeast and demonstrated the importance of water supply stress using projections by 2050. They showed increase in runoff and sediment yield due to increase in erosivity and/or vegetation cover loss. They also stated that climate change and possible future stressor like population growth, land use change, energy security, and policy shift would interact with surface and groundwater availability (Sun 2013).
The Southeast in past has experienced political tensions over water resources (Manuel 2008). Projected warm weather will increase ET, leading to reduced water availability and ground water recharge (Ingram, Dow et al. 2013; Sun 2013; Walsh, Wuebbles et al. 2014). Uptake of soil water by forests is expected to increase, leading to decline in water yield under increased temperature and decreased precipitation projections (Ingram, Dow et al. 2013). Projected population growth and land use change will worsen the situation and pose threat on the economy and unique ecosystems; and land use change in Southeast which ultimately exacerbates the water scarcity, is faster than any other areas in US (Carter, Jones et al. 2014). Future trends derived from projections for 2010-2060, show 5 to 6.5 % decrease in the net water yield for the western part of the Southeast region including Alabama (Sun 2013). Projected temperature extremes are noticeable in the Southeast region. Projections show 4.3℃ and 7.72℃ rise by mid-century (2036-2065) and late-century (2071-2100) under RCP 8.5 respectively (Vose, Easterling et al. 2017). The historical changes for temperature extremes, however, showed insignificant warming of 0.6℃ (difference between mean of ,1986-2016 and 1901-1960) for the Southeast, which was the smallest increase over the continental US (Meehl, Arblaster et al. 2012; Vose, Easterling et al. 2017). Projections under RCP 8.5 by mid-century also reveal 40-50 days per year with temperature greater than 32℃ as a key temperature threshold (Vose, Easterling et al. 2017). For changes in numbers of nights below 0℃, projections (2041-2070 compared to 1971-2000) show increase of 10-15 days for most the region and more than 20 for the northern part of the Southeast (Katz, Parlange et al. 2003). Southeast is categorized as the second vulnerable to weather and climate disasters in the US for the past three decades (1980-2012); Hurricanes can be considered as disasters for the coastal area and tornados and storms are disasters for inland regions where they are close to the Golf and Atlantic coasts (Carter, Jones et al. 2014). Many factors contribute to the climate of the Southeast region including closeness to the Atlantic Ocean and the Gulf of Mexico and El Nino-Southern Oscillation (ENSO), and land falling tropical weather systems (Katz, Parlange et al. 2003). There are few studies that specifically addressed the effects of climate change on the southeast. On the other hand, the scope of studies was the entire Southeast of US and land use change had not been considered to quantify the rapid land use change in the area (Trail, Tsimpidi et al. 2013). Natural hazards like drought, flood, and in general vulnerabilities produced by climate, are results of regional behavior not global (Mahmood, Pielke Sr et al. 2010). Also, it is essential to improve regional projections to determine the mechanisms of the regional forcings and related climate impacts clearly (NRC and CRC 2005). Although, most studies on hydrological future projections have come to the conclusion that water balance components including water yield, surface runoff, ET, baseflow will be affected under in the future under climate scenarios, the hydrological response itself varies depending on region-specific characteristics, topography geography location, and precipitation regimes (Sunde, He et al. 2017). Previous studies have revealed the general impacts of the climate change for the entire Southeast of the US. But they have not used the Coupled Model Intercomparison Projects -Phase 5 (CMIP5) downscaled data with increased robustness and detailed outcome than the Coupled Model Intercomparison Projects -Phase 3 (CMIP3) combined with land use projections. Therefore, in this study, we carried out the investigation on the subbasin scale and with new developed techniques for data preparation. This study provides greater details for a better understanding of the hydrological process, leading in sustainable climate change adaptation. We use SWAT for hydrological modeling and couple it with three representative GCM models in which the data are downscaled using a new developed method called Localized Constructed Analogs (LOCA). In the recent past, researcher have used the same methodology, but with different downscaling method and with predecessors of CMIP5 (Sunde, He et al. 2017; Chen, Marek et al. 2019; Hoyos, Correa-Metrio et al. 2019; Pandey, Khare et al. 2019). Pandey et.al. (2019) have found decline in blue and green water under both RCP4.5 and RCP8.5 pathways for their study area. Comparing different GCMs coupled by SWAT, Sunde et al. (2017) showed decrease in stream flow and increase in ET for their case study. Chen et al. (2017) have investigated land use change and projected 12% to 20 % decline in crop ET by mid and end of 21st century respectively (Chen, Ale et al. 2017). Impacts of afforestation and deforestation on hydrological response have also been studied (McNulty, Caldwell et al. 2013; Sunde, He et al. 2017; Cecílio, Pimentel et al. 2019). Consistent results have not been an outcome of the impact of forest on water yield (Cecílio, Pimentel et al. 2019). Geographical location of the afforestation in Brazilian Atlantic Rainforest showed small significance regarding impact on the average stream flow; however, the minimum flow was reduced (Cecílio, Pimentel et al. 2019). Longer growing season and increased wildfire likelihood and reduction in stream flow were projected for southeast US forests (McNulty, Caldwell et al. 2013). Hoyos et al. (2019) have investigated the hydrology response to drought and reported that the watershed feedback relies on climatic mechanisms and catchment characteristics. Since climate variables should be calculated to investigate the climate change impact on hydrology of a given watershed, it is necessary to couple improved downscaled GCMs with SWAT (Pandey, Khare et al. 2019). On the other hand, hydrological cycle response is unique for each watershed due to the different factors engaged in the process. Thus, it is important to study hydrologic behavior in a smaller scale with greater details. Therefore, herein, we study the hydrological response of Upper Choctawhatchee Subbasin (UCS) to address the possible hydrological response under different climate and land use scenarios. The goals of the study are i) to establish a robust hydrological model for the UCS, ii) to couple the detailed projections with a new downscaling method and iii) to analyze the response of the UCS regarding future stressors.