Climate change impacts on lake surface water temperature (LSWT) have been mostly investigated in deep northern lakes, and are less understood in southern hemisphere shallow lakes. We evaluated the seasonal warming rates of a large (surface area c.a. 10000 km²) shallow choked lagoon in southern Brazil, with a 22 yr time series of MODIS-derived LSWT, and meteorological data. We found high LSWT warming, with a rate of 0.6°C dec^-1 in spring and of 0.3°C dec^-1 in summer. We also found a high correlation between water and mean air temperature trends, as well as a substantial shortening of the cold season. Spatially, there was some homogeneity in the warming rates but prominent point spatial differences, which may result from the variability of the tributaries’ temperature or discharge or decreased water transparency. The high warming rates found here are comparable to those found in deep northern lakes although the changes and processes of heating differ. The stronger warming in early spring can result in accelerated process rates and an earlier start of the phytoplankton growing season.

How to cite: Tavares, M. H., Cardoso, M. A., Motta-Marques, D., and Fragoso Jr, C. R.: Warming rates in a large subtropical shallow Brazilian lagoon, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-511, https://doi.org/10.5194/egusphere-egu23-511, 2023.

Design and risk assessment of large hydraulic structures, whose failure can cause catastrophic damage to the environment, ecology, life and property, are based on Probable Maximum Precipitation (PMP). It is deemed as the theoretical upper bound of the maximum precipitation that is physically possible over a given area for a specified duration. The conventional approaches for estimating PMP are based on stationarity assumption, i.e., the current climatic conditions will remain unchanged even in the future. But the recent increase in extreme precipitation events across the globe and projected increase in the same for future climatic scenarios raises questions on the validity of the stationarity assumption. This issue has gained attention in recent years, and efforts have been directed towards improving existing approaches and devising novel methodologies that would yield reliable PMP estimates in changing climatic conditions. Several researchers have proposed different variations of the widely used storm maximization method to account for non-stationarities arising due to changing climate. The variations imbibe the potential change in PMP resulting either from the trend in precipitable water or from the complex interaction of drivers of PMP. There is a need to compare their relative performance to quantify if the improvement offered by complex variants is significant compared to simple variants in different parts of the globe. In this study, it is investigated through a case study on the frequent flood-prone Mahanadi River Basin in India. For this analysis, future projections of various atmospheric variables (e.g., precipitation, dew point temperature, precipitable water) were obtained from 5 GCMs (General Circulation Models) corresponding to two CMIP6 SSP (Coupled Model Intercomparison Project-6 Shared Socioeconomic Pathways) forcing scenarios namely, SSP1-2.6 and SSP5-8.5. The PMP estimates obtained from improved variants were also compared with their conventional stationary counterpart to assess the effect of dispensing the stationarity assumption.

How to cite: Bhatt, J., Srinivas, V. V., Srinivas, V. V., and Srinivas, V. V.: Comparison of different variants of storm maximization method for Probable Maximum Precipitation estimation in changing climate, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-453, https://doi.org/10.5194/egusphere-egu23-453, 2023.

Extreme events are becoming more frequent, intense, and prolonged, with significant impacts on human lives, the environment, and regional development. Rising temperature due to global warming is one the crucial factors causing the rise in the frequency and severity of extreme events. The worst extreme events are caused by a combination of more than one factor/variable, which makes it essential to study the co-occurrences of extremes, also known as compound extremes. In the present study, four compound extremes are assessed over India for the time period of 1971-2020 using two different statistical approaches, empirical approach, and multivariate distribution analysis. The variables used are temperature and precipitation, having a resolution of 0.25° × 0.25°. Firstly, the empirical analysis of four compound extremes, Hot-Dry, Hold-Wet, Cold-Dry, and Cold-Wet, is done by providing threshold percentiles (25^th and 75^th) to count the exceedance, and Mann Kendall's trend, p-value, and Sen's slope are calculated to assess the changes and significance of the extremes. Next, multivariate distribution analysis using Copula is conducted to study the dependency between temperature and precipitation. The results indicate a significant increase in compound Hot-Dry and Hot-Wet extremes across the country, with a decrease in Cold-Dry and Cold-Wet conditions. The changes in extremes are more pronounced in the later period (1996-2020) than in the earlier period (1971-1995). This study provides insight into the evaluation of compound extremes in India over the past 50 years, which can help us understand the changes and frequency of their occurrence.

How to cite: Sahoo, T., Bejagam, V., and Sharma, A.: Assessment of compound extremes using statistical methods in India, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2305, https://doi.org/10.5194/egusphere-egu23-2305, 2023.

Annual Maxima (AM) and Peaks over Threshold (POT) are the two most common approaches to define extreme time series in hydroclimatic variables. Both methods present limitations. AM frequently fails to include significant extremes that occur during the same year. Conversely, POT may only include clustered values from a few years thus excluding many years from the analysis, especially when the threshold is set high. Additionally, a big challenge in POT is identifying the threshold which can markedly affect the results.

Here, we merge notions from both AM and POT, preserving the strengths of each approach, to extract extreme temperature series and estimate the trends in their frequency and magnitude. We select the values larger than or equal to the minimum of the AM series as high temperatures (HT) (lower than or equal to the maximum of the Annual Minima as the low temperatures – LT). Thus, each year of the HT, LT series has at least one extreme value (H1, L1). We apply the method to 4797 quality-controlled raw station observations from a global dataset of maximum and minimum temperatures over 1970-2019 when warming accelerates. To examine changes in H1-L1 frequency and magnitude, we estimate the ratio of observed to expected H1 (L1) annual occurrences, and the difference between the observed and expected mean H1 (L1) annual temperature values, respectively. We estimate the regression slopes of these ratios at the station level, regionally in 2°×2° grids, and globally. We then compare these trends with the ones obtained from AM and POT series. The proposed method adapts the threshold for each sample, and finds a compromise among all tested methods, thus being a flexible approach that can be applied to any non-intermittent variable.

Acknowledgment

This research was supported by a GWF Ph.D. Excellence Scholarship from the Global Institute for Water Security (GIWS), University of Saskatchewan

How to cite: Nerantzaki, S., Papalexiou, S. M., Rajulapati, ‪. R., and Clark, M.: Estimation of global extreme temperature trends by merging Annual Maxima and Peaks Over Threshold, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12596, https://doi.org/10.5194/egusphere-egu23-12596, 2023.

In combination with Sen's slope, the non-parametric Mann–Kendall (MK) test is one of the most often used statistical techniques for determining a time series' trends. A serially uncorrelated time series is required for the MK test since the autocorrelation in the dataset seriously affects the type 1 and type 2 errors and reduces the performance of the MK test in detecting the statically significant trend. To mitigate this problem, numerous prewhitening techniques (PW, PW-Cor, TFPW-Y, TFPW-WS, VCTFPW; See Collaud Coen et al., 2020) have been developed that effectively reduce lag-1 autocorrelation. In this work, we have proposed a new prewhitening scheme (named as TFPW-Mod) and compared it with previous prewhitening schemes by constructing 5000 linear-trend superimposed (β) AR1 time series with lag-1 autocorrelation (ρ₁) using Monte Carlo simulation. We found that the new prewhitening approach keeps a very good balance between maintaining a low number of type 1 and type 2 errors. The results show that the occurrence of both types of errors largely depends on the length of the time series, with longer periods leading to a strong reduction of errors and to lower bias in the trend slope estimation. For weaker trends and/or the low number of samples, TFPW-Mod couldn’t restore the power of test. However, for a strong trend, this method yields the strongest power, almost independent of the lag-1 autocorrelation. The slope estimation of TFPW-Mod is robust for lower/Medium ρ₁, but significantly deviates from the original trend for highly correlated time series. In most cases, β_TFPW-Mod has lower RMSEs than β_VCTFPW, and leads to the unbiased slope estimation with better accuracy.

How to cite: Sheoran, R. and Dumka, U. C.: A new prewhitening approach for trend analysis in the autocorrelated time series., EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6544, https://doi.org/10.5194/egusphere-egu23-6544, 2023.

In the regions with temperate climate, solid precipitation usually prevails during the winter time. However, a warming climate could alter the timing of snow accumulation and resulted amounts with major impact on the hydrological cycle. This study analyses the changes in the monthly snow-to-liquid precipitation ratio (SLPR)  over the October-May interval in Romania, based on daily precipitation, air temperature and snow depth data provided by 114 weather stations from the national meteorological monitoring network, over the 1961-2021 period. The observed trends showed a country-wide and significant decline in SLPR. The most notable decline is observed during the late winter and early spring months (February-March), with decreasing trends at over 70% of the weather stations, although only 20% suggest statistically significant changes (p value < 0.05). The autumn months (October and November) depict no statistically trends. The trends observed in the late spring (April and May), show a strong decline in SLPR for most mountain weather stations (above 1,000 m), at rates that could exceed 5% per decade (e.g., Tarcu weather station, 2,200 m, the Southern Carpathians). Evidence of elevation dependency of SLPR trends has been found in spring. The results show that the SLPR declines with altitude, especially in April (R² = .30) and May (R² = .67), when the correlations are statistically significant (p<0.05).

This work was co-funded by the European Social Fund, through Operational Programme Human Capital 2014-2020, project number POCU/993/6/13/153322, project title  “Educational and training support for PhD students and young researchers in preparation for insertion into the labor market”.

How to cite: Amihăesei, V.-A., Micu, D. M., Dumitrescu, A., Cheval, S., Bîrsan, M.-V., and Sfîcă, L.: Observed trends in the snow-to-liquid precipitation ratio over Romania, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-10305, https://doi.org/10.5194/egusphere-egu23-10305, 2023.

Natural disasters like droughts and flood events have frequently been occurring due to climate change in the global pattern of precipitation in recent decades. Estimating the future spatiotemporal precipitation variability is necessary to mitigate climate change's impact, particularly extreme precipitation events. However, global and regional climate models typically vary on the projected change in precipitation characteristics over particular regions. Therefore, this study comprehensively evaluates historical and future climate models in terms of spatial distribution, annual cycles, and frequency distributions of precipitation over the Blue Nile basin (BNB) based on different statistical indices. Also, the autocorrelated time series data were subjected to the Mann-Kendall (MK) and Sen's slope estimator tests to identify trends. Many regional and global climate models, such as HadCM3, ECHAM5, MPI-ESM-LR, and EC-EARTH, are employed not only better to understand the discrepancy and the uncertainties of climate models but also to estimate the impact of climate change in the extreme precipitation events over the Blue Nile basin (BNB). Overall, our finding would serve as a benchmark for flood risk mitigation research and water resources management applications over the Blue River basin.

How to cite: Abdelmoneim, H., Kantoush, S. A., Saber, M., Moghazy, H. M., and Sumi, T.: Exploring and investigating the performance of the global and regional climate models in precipitation over The Nile River basin, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-8992, https://doi.org/10.5194/egusphere-egu23-8992, 2023.

In order to effectively plan, design, and manage water resources, it is necessary to understand the trends present in hydro-climatic variables such as streamflow and rainfall. In this study we used the Pettitt’s test as well as the standard normal homogeneity test (SNHT) to discover the trends in streamflow in the Upper Narmada Basin during the 1990 to 2018 period. The Upper Narmada basin extends over an area of 45, 580 square kilometers lies between latitudes 21°20’ N and 23°45' N and longitudes 72°32' E and 81°45’ E in India. From the flow records from gauges in this study basin, change points in the flow regime are thus identified.

Additionally, we performed Mann–Kendall (MK) test, modified Mann–Kendall (MMK) test, Sen's slope (SS) analysis to quantify the trends in streamflow time series. While the MK and MMK tests determine whether a trend is monotonically increasing or decreasing over time, SS suggests the rate of temporal change of streamflow variable. Further, we used advanced machine learning algorithms such as random forest (RF) and long short-term memory (LSTM) to develop flow forecasting models for few gauging sites in the study basin. In this way it is possible to address gaps in the flow records and perform long term analysis of gauge data.

Keywords: Trend analysis, Change point detection, Machine Learning Algorithm, LSTM, Upper Narmada Basin

How to cite: Barbhuiya, S., Ramadas, M., Jena, S., and Biswal, S.: Trend Analysis and Forecasting of Streamflow in the Upper Narmada Basin using Random Forest (RF) and Long Short-Term Memory (LSTM) Models, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-10952, https://doi.org/10.5194/egusphere-egu23-10952, 2023.

The global hydrological cycle is changing in response to climate change and anthropogenic influence. The rainfall, on an annual or sub-decadal timescale, has exhibited erratic and substantial deviation from the long-term average in different parts of the world. Consequently, there are epochs of higher or lower than average rainfall but these are missed in long-term monotonic trend.

As commonly experienced, the rainfall within different geographical regions also varies significantly in terms of magnitude and timing, and on a smaller spatial scale. Integrating large geographical areas and long timescales for monotonic rainfall trend analyses for the meteorologically homogenous regions provides a general picture which is useful for the purpose of administration, water management and distribution. However, it cannot discern the decadal to multi-decadal rainfall variation in different parts of a meteorologically homogenous region and hence a more advanced and comprehensive approach is required for advancing scientific understanding about the hydrometeorological processes and factors governing multi-decadal rainfall variation.

In this study, an innovative approach is presented which involves: (1) 31 years moving average of percentage departure of seasonal rainfall for 120 years at district level; (2) 15 year sliding slope analyses to identify the year of inflection point based on change in direction of slope; (3) K-Means cluster analyses; (4) normality test of clusters based on Z score; and determination of timeframe during which rainfall trend changed significantly.

This approach was tested in the North East India where the rainfall is derived by one of the most dynamic and complicated meteorological systems. Improving understanding about rainfall variability in Northeast India is also very important from ecological, environmental and strategic point of views. Using the above approach, long-term rainfall data (1901-2020) has been analyzed at a district level in Northeast India. Using K means clustering method time windows of prominent change in rainfall trend have been identified. It is inferred from this study that Northeast India has experienced three major climatological events, during 1929-1941, 1961-1971, and 1984-1992. The first and the third events involving a trend reversal from increasing to decreasing nature around 1929-1941 and 1984-1992 affected respectively 49% and 43% area of Northeast India. The second event involving a trend reversal from decreasing to an increasing trend around 1961-1971 impacted 38% of northeastern India.

The meteorological processes corresponding to these timeframes, which could have caused these major rainfall trend reversals are being examined.

How to cite: Chakra, S., Oza, H., Ganguly, A., Padhya, V., Pendey, A., and Deshpande, R. D.: An innovative approach to discern variation in long-term regional monsoonal rainfall trend in North Eastern India., EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11523, https://doi.org/10.5194/egusphere-egu23-11523, 2023.

Both climate change and intensified urban development will affect the temporal and spatial distribution of rainfall. There is still a lack of relatively quantitative and comparative studies. Most cities in China entered the rapid urbanization stage around 2000. This study focuses on the megacities Beijing, Shenzhen and Hong Kong. Based on rainfall observation data and land use and other remote sensing image data, this study investigates the impacts of the impervious areas on the temporal and spatial changes of short-duration and long-duration heavy rainfall before and after rapid urbanization. The impact of climate change on rainfall intensity is also analyzed. The results show that there are huge spatial shifts in rainfall after rapid urbanization. For example, it is observed that the center of the heavy rainfall shifts from northeast to southwest, and the hydrological homogenous areas in Beijing have increased from two to three after rapid urbanization. Meanwhile, with the influence of climate change, the intensity of rainfall is increasing, but due to urbanization, the increase varies greatly in each hydrological homogenous area. Shenzhen and Hongkong, the neighbor cities without physical boundaries, have entered into different urbanization stages since 1990. Hong Kong has developed very slowly, while Shenzhen has developed very rapidly. As a result, the short-duration heavy rainfall is mainly concentrated in Shenzhen, and the long-duration heavy rainfall is mainly concentrated in Hongkong. The short-duration heavy rainfall in Shenzhen mostly occurs at noon, while that in Hong Kong mostly occurs in the morning. The urbanizations of Hong Kong and Shenzhen are different, which resulted in different impacts on rainfall in these two cities. This study examines the impact mechanisms of urbanization and climate change on rainfall in cities in different climate zones. It provides scientific basis in supporting the city planning and the development of resilient cities.

How to cite: Huang, J. J.: The study of the changes in precipitation induced by intensified urbanization and climate change by observational records, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-10274, https://doi.org/10.5194/egusphere-egu23-10274, 2023.