Spatio-temporal Evolution of Wet-dry Event Features and their Transition across Upper Jhelum Basin (UJB)-South Asia

The increasing rate of occurrence of extreme events (Droughts droughts and floods) and their rapid transition magnifies the associated socio-economic impacts than with respect to those caused by the individual event. Understanding of spatio-temporal evolution of wet-dry events collectively, their characteristics and transition (wet to dry and dry to wet) is therefore significant to identify and locate most vulnerable hotspots, providing the basis 10 for the adaptation and mitigation measures. The Upper Jhelum Basin (UJB)-South Asia was selected as a case study, where the relevance of wet-dry events and their transition have not been assessed yet, despite of clear evidence of climate change in the region. The Standardized Precipitation Evapotranspiration Index (SPEI) at the monthly time scale was applied to detect and characterize wet and dry events for the period 1981-2014. The results of temporal variations of SPEI showed a strong change in basin climatic features associated with El Niño Southern 15 Oscillation (ENSO) at the end of 1997, with the prevalence of wet and dry events before and after 1997 respectively. The results of spatial analysis show a higher susceptibility of the monsoon-dominated region towards wet events, with more intense events occurring in the eastern part, whereas a higher severity and duration is featuring in the southwestern part of the basin. In contrast, westerlies dominated region was found to be the hotspot of dry events with higher duration, severity, and intensity. Moreover, the surrounding region of the Himalaya 20 divide line and the monsoon-dominated part of the basin were found to be the hotspots of rapid wet-dry transition events.


Introduction
There is growing evidence that recent warming is leading to significant alteration in hydrological cycle, exacerbating extreme weather events in general (Peterson et al., 2012) in many regions of the world. Extremes 25 weather events such as floods and droughts and their rapid successions (recurrent spells) during past few decades have taken a heavy toll on both life and property. Moreover, such events can have large impacts on water availability, agriculture and food security, power production, and natural ecosystems (He et al., 2019, Sheffield and. These events are projected to regionally intensify and be more frequent within the context of global warming, underscoring the importance of research on wet-dry extremes weather events collectively. The 30 climate change projections for Asia continent in the fifth sixth Assessment Report (AR5AR6) of Intergovernmental Panel on Climate Change (IPCC) reported that during the 21 st century South Asia is likely to face more intense and frequent heatwaves and humid heat stress, whereas both annual and summer monsoon precipitation will increase, with enhanced inter-annual variability (medium confidence) (Zhongming et al. 2021).
warming in South Asia region is likely to be above global mean and climate change will im pact the glacier and

Data description
The monthly temperature and precipitation data from 1981-2014 were used in this study. Due to the spatially 145 limited coverage of observed climatic stations over high altitudes in the basin (see Figure 1), the distribution mapping (DM) based corrected ERA5 estimates (0.25 o horizontal resolution) were used. ERA5 is a relatively new reanalysis launched by European Centre for Medium-Range Weather Forecasts (ECMWF) (Saha et al., 2010).
The data are developed by using advanced 4Dvar assimilation scheme and provide various atmospheric variables at 139 pressure levels for 1979-present time. The suitability of ERA5 precipitation and bias correction method 150 with respect to extreme precipitation analysis has been checked against observed station data in previous works (Ansari and Grossi, 2021). The observed temperature data from 20 climatic stations located within and adjacent to UJB were used in deriving the potential evapotranspiration (PET) using the Thornthwaite equation (Thornthwaite, 1948). PET values were interpolated at 0.25 o using Kriging with External Drift (KED) with consideration of elevation as a predictor (Goovaerts, 2000).

155
The daily observed precipitation and temperature data of 15 climatic stations located within the political boundary of Pakistan were collected from Pakistan Meteorological Department (PMD) and Water and Power Development Authority (WAPDA). For the Indian side region, Indian Meteorological Department (IMD) daily gridded precipitation and temperature datasets, derived from a dense network of meteorological stations for the Indian mainland (Pai et al., 2015), were extracted at five stations and used for that region. The analysis was carried 160 out for a period of 34 years , due to the availability of observed data. In fact there are only a few climatic stations where data are available starting from 1971, but the number of stations would not be enough for the spatial analysis. The observed temperature data was used to calculate potential evapotranspiration (PET) using the Thornthwaite equation (Thornthwaite, 1948) due to data limitation. A study conducted by Beguería et al. Commented [RA8]: RC1: As the data use to carry out a research work is the base of a research work and the most important ingredient. The authors have not provided any detail of the data they have used to carry out their work. I suggest that the authors must provide the complete detail of the data they have used in this research work. Moreover, the authors have applied any homogeneity test on the data to ensure the data quality? In data description section the authors did not mention from where they took observed data and what is the ethnicity of the data. I suggest the authors to go through the Zaman et al 2020 for the data quality and presentation.

RC2
: I see several problems with data. First, I can't read the source of observed temperatures. Then, the reliability of ERA5 precipitation data needs to be accurately checked against available observations. In this regard, the authors provide a reference to a conference abstract (Ansari and Grossi, 2021). It's not enough, a section about data validation is needed. Finally, I'm not that keen on using the Thornthwaite method, which is very dated. I would suggest using at least a temperature-based model, e.g. Hargreaves-Samani. However, ERA5 provides potential evaporation data, a comparison between such data and the results achieved by the authors with another method would be interesting and could provide useful insights. The authors should discuss their choice of relying partially on datasets and partially on ground observations.  reconstruction at all temporal resolutions and station densities, but its influence was less clear for daily to monthly precipitation. Furthermore, all spatial interpolation techniques can perform poorly in regions with insufficient high-elevation data, due to inaccurate estimation of local lapse rates (Ruelland and Sciences, 2020). Therefore, 175 the distribution mapping (DM)-corrected ERA5 precipitation estimates (0.25o horizontal resolution) were used in the present study. ERA5 is a relatively new reanalysis launched by European Centre for Medium-Range Weather Forecasts (ECMWF) (Saha et al., 2010). The data are developed by using advanced 4Dvar assimilation scheme and provide various atmospheric variables at 139 pressure levels for the period 1979-present time. The DM method adjusts the cumulative distribution function (CDF) of modelled precipitation to match with the observed 180 precipitation CDF using a transfer function (Sennikovs and Bethers, 2009) and it is commonly used to correct the systematic distributional biases (Cannon et al., 2015). The Gamma distribution (Thom, 1958) with a shape and a scale parameter was found to be suitable for the precipitation distribution in the study region (Azmat et al., 2018).
The suitability of ERA5 precipitation and bias correction method with respect to extreme precipitation analysis was checked against observed station data and a few results of the reliability check of DM-corrected ERA5 is 185 provided in supplementary material (see figure S1).

Wet and Dry Events Identification
We adopted SPEI, a most widely used index, SPEI, was adopted to detect and characterize wet and dry events of different severity levels (moderate, severe, and extreme). The SPEI index allows support comparisons over 190 time and space, as proxies of wet and dry conditions from both the meteorological and agricultural perspectives.
Although the SPEI was originally proposed for drought monitoring, it can also be used as a tool to detect flood risk. The calculation procedure of SPEI involves two steps: fitting a log-logistic distribution to the monthly climatic water balance (P-PET) time series and then transforming the cumulative probability of the fitted distribution to a standard normal distribution (with mean zero and variance one). According to this distribution 195 method, the probability distribution function of a variable x is expressed as: Where α, β and γ are the shape, scale, and origin parameters, respectively. In the second step, SPEI is calculated as the standardized values of F(x) as follows: Where Commented [RA11]: RC1: Line 138, I would strongly suggest adding 2-3 sentences why authors prefer to use distribution mapping method of bias correction of ERA5 precipitation and which frequency distribution was employed/fitted to the precipitation data.  Table 1. Positive and negative value of SPEI represent the severity of wet and dry events, respectively.

275
Moreover, the floods and flash droughts are not clearly associated with long term SPEI, because the averaging effect of long-term accumulated precipitation and temperature surpassed surpasses the signal of extreme precipitation and temperature over a short period. Flash drought is relatively a new type of drought. Currently, there is not a universally accepted definition or criteria for flash drought, though there is general consensus on the principle of rapid onset or intensification characterized by moisture deficits and abnormally high temperatures 280 for a period lasting at least 3 weeks (Lisonbee et al. 2021, Otkin et al. 2018, Hunt et al. 2009). This highlights the usefulness of SPEI at the monthly scale in representing flood and flash drought events. It is noted that the terms "wet-dry events" or "wet-dry months" presents similar meaning for our study, as the analysis was made at the monthly time step. A clearer picture of the monthly evolution of wet/dry events of different severity levels and their seesaw variability can be seen in Table 2. The SPEI-1 values fluctuate remarkably from one month to another.

285
For example, an extreme wet October in 1987 was followed by a severe dry November, and a severe wet June occurred at the tail of the longest drought spell in May 2001. Such rapid transition from wet-to-dry and from dry-to-wet events were was more prominent during the first half of the study period (before the year 1997).

310
The wet/dry event characteristics (duration, severity, and intensity) were computed for each pixel to analyze their spatial distribution. Pixel based analysis shows the location of the most vulnerable parts of the basin, providing the basis for future decisions onthe adaptation and mitigation measures. In this study, the total, average and maximum value of duration, severity, and intensity were computed for the study period . The maps of wet and dry duration are displayed in figure 4. Overall, the study area encountered relatively more wet 315 months than dry months during the whole study period. The total wet duration (TWD) and total dry duration  The spatial distribution of total, average and maximum severity of wet/ dry events are is presented in figure 5. All wet/ dry severity maps show similar spatial patterns as wet/dry duration maps. In terms of total wet

Wet-Dry Transition Time
The number of transitions and their average transition time for wet-to-dry and dry-to-wet events for the period 1981-2014 is presented in figure 8 and figure 9. As expected, the number of transitions for wet-to-dry and 385 dry-to-wet event was the highest for the moderate level of events, followed by severe and extreme levels of events.
Consequently the average transition time from wet-to-dry and dry-to-wet event was found to be the highest for the extreme level of event followed by severe and moderate levels of events. The number of transitions for moderate, severe, and extreme levels of events varies from 15 to 26, from 6 to 16, and from 1 to 5 respectively.
Overall, the number of transitions for dry-to-wet event is larger than the wet-to-dry event for severe and extreme 390 levels of events, whereas the opposite was found for the moderate level of events. The average transition time intervals (months) from wet-to-dry and dry-to-wet event for the period 1981-2014 is presented in figure 8. As expected, the transition time from wet-to-dry and dry-to-wet event is highest for the extreme level of event followed by severe and moderate levels of events. The transition time for moderate, severe, and extreme levels of events varies from 1.8-to 6.5, from 1.8-to 16.75 and from 3.5-to 187.0 months, respectively. Overall, 53.57% 395 and 17.86% of pixels in the UJBa greater number of pixels showed longer transition time from wet-to-dry than from dry-to-wet for moderate and extreme levels of events, whereas the opposite was seen for the severe level of events.

Wet/Dry Rapid Transition Events
The wet/dry rapid transition is the consecutive occurrence of wet and dry months of any severity level.
The frequency of wet-to-dry (wet month is followed by dry month) and dry-to-wet (dry month followed by wet 410 month) rapid transition events were computed for each grid cell and are shown in figure 910. The frequency of wet/dry transition events varies/ranges from 5 to 20 events during 34 years of study period. About 50% pixels in the UJB encountered more number of wet events terminated at dry months. In general, the basin encountered a greater number of wet events terminated at dry months. The spatial distribution of frequency of wet/dry rapid transition events revealed that the wet-to-dry events are less frequent over the westerlies dominated region of the 415 basin, whereas the southwestern part of the basin was more affected by the wet-to-dry abrupt altered events. By contrast, dry-to-wet abrupt altered events are found to be more frequent over pixels surrounding the Himalaya divide line, whereas the remaining part of the basin depicts less incidence of dry-to-wet altered events.