Linking Future Hydroclimatological Changes With Past Climatic Conditions In Southeastern Iran: Insights From Models And Observations

We compare the predicted results of future hydrological changes based on a thirty-year (1989-2019) weather dataset with paleoclimatic changes inferred based on established proxies from the Jazmurian playa in southeastern Iran. Parallels between expected changes in the future were compared to past climatic conditions to trace the impact this region has undergone in the distant past. The study area is affected by the Indian Ocean Summer Monsoon (IOSM) and the Mid-Latitude Westerlies (MLW). The maximum and minimum temperatures and precipitation were predicted for the future (2061- 2080) by statistical downscaling outputs of 5 GCM models (EC-EARTH, GFDL-CM3, HadGEM2-ES, MIROC5, MPI-ESM-MR) under RCP 4.5 and RCP 8.5. The results show that the 20-years average of the mean temperatures ((T max + T min )/2) will increase in the range of 3.2 to 4.6 °C under RCP 8.5 compared to the base period. The trends suggest that the region will experience drier conditions than the baseline period in the future under both scenarios. In addition, the GCM predicts a considerable decline in MLW precipitation and little change in future IOSM precipitation under both scenarios compared to the baseline. The decrease in MLW precipitation is consistent with other GCM predictions and real paleoclimatic changes that happened during past warm/wet periods in the region. However, considering the close relationship between the increase in the Earth’s radiation budget and enhanced IOSM precipitation in southeast Iran since the late Pleistocene, we postulate that more intensive IOSM activity can be expected in the future. changes based on different global warming scenarios are used to evaluate the performance of the models. To verify if these changes may have also occurred in the past when no direct measurements of precipitation or temperature are available, we refer to the paleoclimate study by Vaezi et al. (2019) on a sediment core from the Jazmurian playa. This study reconstructs the linkages between paleoenvironmental conditions and variability in IOSM and MLW outputs that contributed to various environmental changes in the interiors of West Asia since the late Pleistocene. We compared the variations in future atmospheric circulation and related changes in precipitation to the past cold and warm periods. The predicted simulations and paleoclimate events superimpose a complex mosaic, which can gauge our response and adaptation to climate change scenarios.


Introduction
Earth's climate has varied due to natural causes such as the change in solar insolation, orbital motion, orogeny, and ocean circulation pattern in the past (Bytnerowicz et al., 2007;Nakicenovic et al., 2000). However, since the industrial revolution began during the 18th -century, a sharp increase in greenhouse gas emissions (i.e., carbon dioxide, methane, nitrous oxide, and ozone) have resulted in global warming and a drastic change in existing climatic conditions (Le Treut et al., 2010; IPCC 2019;). In particular, human activities associated with the burning of fossil fuels have increased carbon dioxide levels in the atmosphere from 280 ppm in 1750 to more than 400 ppm in recent years (IPCC 2019(IPCC , 2021. Since the 1970s, globally averaged surface temperature data have shown a linear warming trend of ca. 0.9°C (Millar et al., 2017). Unlike temperature, which has increased globally, precipitation records indicate a variable response both in frequency and intensity (Archer & Rahmstorf, 2011;Konapala et al., 2020). Hence, there is an ongoing debate about how global warming will affect future precipitation because it has important implications for agriculture practices and food supply (Donat et al., 2016).
The lives of more than two billion people in South and West Asia depend on the summer monsoon precipitation on both short and long timescales (Clift & Plumb, 2008;Cullen et al., 2000). Therefore, understanding how the monsoons will change in the face of the anticipated increase in greenhouse gas emissions and rising global warming is a fundamental challenge. Also, the General Circulation Models (GCMs) face di culty in simulating the regional distribution of monsoon (Turner & Annamalai, 2012) due to a multitude of physical processes and interactions that in uence precipitation (Sperber et al., 2013). To reduce the uncertainty in climate models associated with monsoons and their intensity, we must understand the processes driving monsoons, seasonality, and uctuations (Turner & Annamalai, 2012; Wang et al., 2017;Zhisheng et al., 2015).
It is suggested that the increase in greenhouse gas concentrations will intensify monsoons mainly due to an increased land-sea difference in temperature and a northward shift of the Inter-Tropical Convergence Zone (ITCZ) (Cao & Zhao, 2020;Li & Ting, 2017;Sachs et al., 2009). One such region of the world is southeastern Iran, which lies on the extreme northern border of the monsoonal domain that may be signi cantly affected by changes in the monsoon pattern and intensity. Southeastern Iran, straddled between the Indian Ocean Summer Monsoon (IOSM) precipitation zone and the Mid-Latitude Westerlies (MLW) precipitation zone, makes it highly sensitive to changes in climatic conditions Rashki et al., 2021;Vaezi et al., 2019). Furthermore, paleoclimate records indicate that intensity and variation of IOSM and MLW have changed signi cantly since the late Pleistocene affecting the regional hydrological conditions (Vaezi et al., 2019;Clift & Plumb, 2008;Stevens et al., 2001). Therefore, establishing a better understanding of atmospheric circulation patterns and precipitation in the distant past could help in improving our assessment of future climate change scenarios and variations in regional precipitation patterns (Mehterian et al., 2017).
GCMs have been widely used to study atmospheric patterns and eventual effects on the global and regional scales. However, output data from GCMs are typically coarse to estimate the hydrological response to climate change on a regional scale. Thus, there is a need to downscale the data from a coarse resolution in GCMs to a 'local' sub-grid-scale, weather station scale (Busuioc, 2008;Wilby et al., 2002), which can be achieved either by statistical or dynamical methods. Amongst the statistical downscaling methods, the Long Ashton Research Station Weather Generator (LARS-WG) has been extensively applied and tested in different climatic regions (Luo & Yu, 2012;Qian et al., 2004;Semenov et al., 2002Semenov et al., , 2013Semenov & Barrow, 1997;Street et al., 2009). The simulations in these studies highlight the capability and accuracy of the model in simulating climate change and projections for the future.
In the present study, as a comprehensive climate-driven investigation of the arid Iranian plateau, we developed a general understanding of the qualitative and quantitative impact of changes in precipitation and temperature pattern and their impacts. In this context, daily precipitation and daily maximum (T max ) and daily minimum (T min ) temperatures in the Jazmurian playa in southeastern Iran were evaluated for the distant future scenario extending from 2061-2080 using statistically downscaled outputs from the 5 GCMs with the LARS-WG model under RCPs 4.5 and 8.5. The predicted results of future hydrological changes based on different global warming scenarios are used to evaluate the performance of the models. To verify if these changes may have also occurred in the past when no direct measurements of precipitation or temperature are available, we refer to the paleoclimate study by Vaezi et al. (2019) on a sediment core from the Jazmurian playa. This study reconstructs the linkages between paleoenvironmental conditions and variability in IOSM and MLW outputs that contributed to various environmental changes in the interiors of West Asia since the late Pleistocene. We compared the variations in future atmospheric circulation and related changes in precipitation to the past cold and warm periods. The predicted simulations and paleoclimate events superimpose a complex mosaic, which can gauge our response and adaptation to climate change scenarios.

Study Area
The Jazmurian playa is an inland basin or depression in southeastern Iran located 350 m above mean sea level. At the center of the Jazmurian basin is a seasonal lake that sporadically lls with fresh water. The playa receives ca. 75% of its annual freshwater input from the Halil River in the west. The Bampur River in the east is the other source of fresh water in the playa (Frs, 1975). The playa lies at the northern margin of the ITCZ (Fig. 1A). Scorching and dry summer and relatively mild winter result in high evaporation (ca. 2500 mm/year; Rashki et al., 2017), resulting in arid conditions. The local temperature and precipitation are reported by the Kahnuj and Iranshar stations located to the west and east of the playa, respectively (Fig. 1B). The mean annual temperature at the Iranshahr station is 27.2°C and the mean maximum and minimum temperatures during the warmest and coldest periods of the year are 44.5°C (July) and 8.4°C (January), respectively (Fig. 1B). The mean annual temperature at the Kahnuj station is 27.3°C, and the maximum and minimum temperatures during the warmest and coldest months of the year are 44.4°C (July) and 8.4°C (January), respectively. The annual precipitation around Jazmurian is about 138 mm as reported in Kahnuj (for a 30years average), whereas Iranshahr reports it as 99 mm. The average monthly precipitation recorded at these two weather stations indicates the highest amount from January to March and dry conditions extending from April to November (Fig. 1B). However, the eastern part of the basin also experiences some precipitation during summer.

Materials And Methods
This study uses historical observation weather data (1989-2019) from the two weather stations in the Jazmurian playa (Kahnuj and Iranshahr) to simulate the future scenario. GCM data from the Coupled Model Intercomparison Project Phase 5 (CMIP5) is used namely. EC-EARTH, GFDL-CM3, HadGEM2-ES, MIROC5, and MPI-ESM-MR, to account for intermodal uncertainities. The output from GCMs is statistically downscaled using LARS-WG for the future scenarios RCPs 4.5 and 8.5 for variables precipitation, maximum (Tmax), and minimum (Tmin) temperatures.

CMIP5 GCMs and RCPs
The speci cations and horizontal resolution (latitude/longitude) of CMIP5 global models used in this study are outlined in Table 1. The RCPs represent radiative forcing, where RCP4.5 represents a scenario with medium emissions (650 ppm CO 2 ) and medium radiative forcing (4.5 W/m 2 and RC8.5 represents a more signi cant radiative forcing and increasing radiative forcing projected at 8.5 W/m 2 (1370 ppm CO 2 ) by 2100 (IPCC, 2014). Table 1 Five global climate models (GCMs) were selected for the LARS-WG simulation in this study.

GCMs
Research

LARS-WG Stochastic Weather Generator
The LARS-WG software version 6 is a stochastic weather generator used to simulate long-term weather data at the weather station scale under baseline and future conditions (Semenov & Barrow, 1997;Semenov & Stratonovitch, 2010). The model applies different statistical tests (t-test and Chi-square) for comparing the data generated with observed data during the baseline period and tests the model's performance (Semenov et al., 2002). Thus, developing weather data by LARS-WG can be divided into three steps -calibration, validation, and generation of simulated weather data.
During the calibration step, LARS-WG calculates statistical parameters for probability distributions of weather variables and their correlation based on the observed daily weather data. The program employs a semi-empirical distribution of observed weather data to simulate its statistical characteristics. The model also generates site-speci c cumulative probability distribution function for climate parameters such as wet and dry days, daily precipitation, minimum and maximum temperature, and solar radiation (Semenov et al., 2002). During validation, the probability distributions of climate variables derived from the calibration processes are used to generate synthetic weather time series of arbitrary length having the same statistical features as the historical weather data ( Table 2 The statistical tests compare the observed data for two weather stations in the Jazmurian playa with simulated data generated using LARS-WG for the daily precipitation and daily maximum (T max ) and daily minimum (T min ) temperatures.

Discussion
The 2015 Paris climate agreement aims to hold the average increase in temperature to well below 2°C, and "pursue efforts" to limit this to 1.5°C. This is the only way to avoid adverse and unpredictable weather effects both on local and The multi-proxy climate record from Jazmurian playa reveals that the regional hydrology of southeastern Iran since ca. 14.7 cal kyr BP is primarily governed by the IOSM strength, which in turn is linked to the position of the ITCZ in response to orbital-scale changes in summer insolation (Fleitmann et al., 2007;Gupta et al., 2003;Overpeck et al., 1996). Since 14.7 cal kyr BP (e.g. Bølling-Allerød and early Holocene epochs), Earth's radiation budget has increased signi cantly on the Indian Ocean and the adjacent Asian landmass resulting in intensi ed IOSM in southeastern Iran (Vaezi et al., 2019). Notably, the position of the ITCZ has also shifted substantially during the past millennium.
Lacustrine records from as far as the Galápagos and Palau Islands show that the ITCZ reached its southernmost position during the Little Ice Age, which lasted from 1400 to 1850 AD, and since then, it has been creeping northward over the past 300 years, possibly due to higher solar irradiance (May, 2002 these results need to be evaluated carefully in that context. Moreover, the results do not match the paleohydrological changes and intensity of IOSM during past warm periods.

Conclusions
In this study on future and paleoclimate changes in an arid region of southeastern Iran situated on the northmost border of IOSM, we presented 1) future simulated precipitation and temperature shifts based on different scenarios in GCMs, and 2) trace real paleoclimatic changes that happened in the Jazmurian playa since the late Pleistocene. The playa's maximum and minimum temperatures and precipitation projections are estimated from the downscaled ve CMIP5 GCMs (EC-EARTH, GFDL-CM3, HadGEM2-ES, MIROC5, and MPI-ESM-MR) under RCPs 4.5 and 8.5 using LARS-WG. The paleoenvironmental records are used to examine the relationship between predicted changes in precipitation (variability in IOSM and MLW output) based on two different scenarios.
Future projections of the 20-years average (2061-2080) of the mean temperatures will increase to 3.2 to 4.6°C under RCP 8.5 in southeastern Iran compared to the baseline period. However, the Iranshahr station shows little change in future precipitation compared to its baseline. HadGEM2-ES under RCP 8.5 showed a slight increase (about 12.2 mm) in monsoon precipitation compared to the baseline condition, and it is the highest increase among the GCMs. On the other hand, GCMs predicted a considerable decrease in MLW precipitation in Kahnuj station. Similarly, HadGEM2-ES and GFDL-CM3 indicate a weak decline in precipitation compared to other GCMs at the Kahnuj station for RCP 8.5.
The current study indicates that in the Jazmurian playa, MLW precipitation will decrease, and temperature will increase in future simulations. Infact, Consistent with several GCM studies and real paleoclimatic changes that happened during past warm/wet periods (e.g., the Bølling-Allerød and the Early Holocene) in the region, results emerging from the Jazmurian playa also show a weakening of the MLW under the RCP 8.5. However, in future simulations, we also noted that IOSM projections in Jazmurian playa do not match the actual shift in IOSM intensity or strength, especially during the warm/wet periods. Nevertheless, the results of the multi-proxy data from Jazmurian playa and other highresolution records represent a similar pattern of intensive IOSM activity and northward movement of the ITCZ in response to the orbital-scale changes in summer insolation.