مقالات رزومه دکتر سیده عاطفه محمدی

In this study, a future projection of four extreme precipitation indices consists of total extreme precipitation (R95p), extreme precipitation days (R95d), the fraction of total rainfall from events exceeding the extreme precipitation threshold (R95pT) and extreme precipitation intensity (AEPI) over Iran was carried out using the reference period of 1990-2014 based on the multi-model ensemble approach and a rank-based weighting method using five models from the CMIP6 models (MPI-ESM1-2-HR, EC-Earth3, EC-Earth3-Veg, GFDL-ESM4, and MRI-ESM2-0). The weight of each model was calculated depending on its historical simulation skill and then weighted and unweighted groups were used for future projections. In this study, the threshold of heavy precipitation was calculated using percentiles, according to the method of Zhai et al. (2005). For this purpose, at each grid point, annual rainfall data (daily rainfall greater than or equal to 1 mm) were sorted in ascending order to obtain the sequence of annual rainfall for each year from 1990 to 2014. Then, the mean 95th percentile for the 25-year precipitation sequence at each grid point was defined as the extreme precipitation threshold (Rwn95). It should be noted that the heavy rainfall threshold for observations and CMIP6 models was calculated separately from observation data of synoptic stations and from each CMIP6 model. Based on this extreme precipitation threshold, four extreme precipitation indices were calculated. The results of the spatial skill of the simulation show that the EC-Earth3 model with MR_taylor equal to 0.65 and with a horizontal resolution of 0.7 degrees has the best skill in simulating the spatial pattern of extreme precipitation indicators, and the MPI-ESM1-2-HR model with MR_taylor equal to 0.5 and with a horizontal resolution of 0.938 degrees is the second suitable model among the five selected models used for simulating the spatial pattern of extreme precipitation indices. Also, the results of the simulation time skill test show the superiority of MPI-ESM1-2-HR and GFDL-ESM4 models with MR_IVS respectively equal to 0.6 and 0.5 compared to other studied models. It is important to note that the horizontal resolution of the model is not the only factor that determines the skill of the model in simulating extreme precipitation indicators in the study area. Because the model with above-average weight (>0.2) is not a high-resolution model and improvement in physical processes which are also needed. The results show that the probability of increasing the total extreme precipitation (R95p) and extreme precipitation intensity (AEPI) in the study area in the period 2026-2050 under four scenarios SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5 with values greater than 0.5 is almost certain. According to the median value close to zero or even negative of the extreme index R95d, it is inferred that the priority of the increase of the extreme amount of precipitation is over the number of days of occurrence, and these extreme precipitations will occur in fewer days, which is a warning for the flood risk. A comparison between the weighted and unweighted ensemble means shows that the uncertainty in the study area is almost always reduced after applying the weighted scheme to future probabilistic projections.

Ultraviolet radiation is defined as electromagnetic radiation with wavelengths in the range of 200-400 nm and is divided into three different bands. UVC is related to the wavelength from 200 to 280 nm, while UVB is related to the wavelength ranging from 280 to 315 nm and UVA is related to the wavelength from 315 nm to the visible level (400 nm). Ultraviolet radiation has beneficial effects such as making vitamin D and disinfecting effects. On the other hand, it causes harm such as burns and skin cancer, and damage to the eyes and immune system. Predicting the amount of UV radiation based on the UV index can be of great help to people's health. In this study, the tropospheric ultraviolet-visible (TUV) model was used to predict the UVI index. This model requires ozone, whiteness, and Aerosol Optical Depth (AOD) to forecast UVI. WACCM model data was used for ozone and whiteness column values from the ozone data of the GFS and AOD global forecasting systems. 612 case studies in the whole year of 2020 were selected from each of the 12 months of the year from different parts of the country. GFS, WACCM, and OMI data were extracted for the mentioned dates and interpolated at the desired points. Because OMI data is available locally at noon everywhere, case studies have been selected for noon. Then the interpolated values along with the length, width, and height of the points were given as input to the TUV model, and the UVI value was predicted. Due to the lack of access to the actual value of UVI in the country, OMI data was assumed as observational data and used to compare with the predicted value. Conventional statistical measures ME, MAE, RMSE, and Pearson correlation coefficient were used to validate the prediction value with observational data. The results showed that in January, February, April, November, and December, which are the coldest months of the year and the day length is shorter and the sun is less intense, so the error rate is lower than in other months (warm months of the year). However, in general, the forecast is very accurate. So that in all selected study cases, the values of ME, MAE, RMSE, and R are 0.16, 0.85, 1.13, and 0.93, respectively, which indicates the high accuracy of the forecast. The results also showed that the forecast error has a linear relationship with the AOD value. Thus, the higher the AOD value, the more negative the forecast error and underestimated forecast value.In the warmer months of the year, the length of the day is longer and the intensity of the sun's radiation is higher, resulting in more errors. The amount of error is also related to the amount of light depth of the particles; the greater the AOD, the greater the error. The correlation coefficient diagram also showed that there is a high correlation between the forecast and observation values. This research is the first research in the field of forecasting the UV index in the country and has had satisfactory results.

Numerical weather prediction (NWP) models are not completely accurate and error free, and there is always some uncertainty. The errors in weather forecasting stem from the limitations of human theoretical understanding of the atmosphere and the operational capacity to produce forecasts. It is necessary to make a forecast, along with an estimate of its uncertainty. This is accomplished by creating ensemble systems of weather forecasts differing in the initial conditions or physical formulation of NWP models. There are several methods for post-processing of ensemble forecasting, including Bayesian Model Averaging (BMA) and Ensemble Model Output Statistics (EMOS) that they are more popular because of higher efficiency and accuracy. In this research, first, an 18-member ensemble system is formed, which each member is an independent run of the WRF model with different physical configurations. BMA method was used to estimate the density function of predicting 24-hour cumulative precipitation. Due to some hardware limitations and access to an ensemble system with fewer number and more efficient members, the size of the ensemble system has been reduced to 7 members. Using the BMA method, a weight is assigned to each ensemble member. The size of the ensemble system is reduced by removing the members who had less weight. The probabilistic prediction verification obtained from the 7-member ensemble system in a test period from 15 January 2020 to 15 May 2020 has been checked using reliability diagram. The results show that the probabilistic predictions are sufficiently skilled for 24-hour cumulative precipitation.

Reanalysis data have been considered as an important source of atmospheric information in a variety of applications such as climate studies, hydrological modeling and numerical weather prediction. Evaluating the effectiveness of the reanalysis products in each area before use is of great importance. With the advent of advanced reanalysis such as ERA5 and ERA5-Land, the interest of many researchers in using these data sources has increased.     To date, several studies have been conducted in the country to statistically compare the reanalysis products with other meteorological data sources, each of which has its limitations and does not provide a comprehensive evaluation of the reanalysis data across the region. In other words, most of these studies are related to the evaluation of one of the meteorological variables such as precipitation and have been done in a specific location of the country or have used a limited number of ground stations in statistical comparison.     In this work, the quality of 2m temperature, surface pressure, 10m wind speed and dew point temperature of ERA5-Land are evaluated temporally and spatially over Iran. For this purpose, sub-daily observations of 406 synoptic stations from 1999 to 2019 were used. The bilinear method was used to spatially interpolate the meteorological values obtained from ERA5-Land at the station locations. After preparing the ERA5-Land sub-daily time series and the corresponding actual observations, the error statistics required to evaluate the ERA5-Land data were calculated. Statistical comparisons between ERA5-Land products and ground observations of 2m temperature, surface pressure, 10m wind speed and dew point temperature parameters are done with a 3-hour temporal resolution.     In the whole region, ERA5-Land products and local measurements of 2m temperature, surface pressure, 10m wind speed and dew point temperature showed agreement about 0.97, 0.98, 0.49 and 0.88, respectively. Also, compared to the actual observations, the mean RMSE for the above ERA5-Land data products achieved 2.87°C, 19.42 hPa, 2.52 m/s and 4.12°C, respectively.     The study of bias values showed that in the region of Iran, ERA5-Land, on average, estimates all the studied variables less than the observed value. In addition, the study of the relationship between height difference of ERA5-Land grid points and station height with estimated error values showed that with increasing altitude difference, the size of negative bias and root mean square error of 2m temperature and the surface pressure of ERA5-Land increases significantly. Therefore, it is suggested eliminating the existing systematic errors in the area before applying this data.

To quantify the uncertainty exists in the weather forecasting, and producing the probabilistic forecasting of diurnal maximum and minimum temperature, two important methods named Ensemble Model Output Statistics (EMOS) and Bayesian Model Averaging (BMA) are used to estimate the forecast probabilistic density function. In this study, the ensemble system has two members consisting of 1 to 5 ahead forecasting of diurnal maximum and minimum temperature over synoptic stations of Iran provinces from 10 November 2017 to 31 May 2018. These members are the outputs of the WRF model with two different physical configurations. Results show that the deterministic post-processed forecasts have improved the root mean squared error (RMSE) of deterministic raw forecast, 37% and 7% for mximum and minimum temperature, respectively. Generally, raw forecasts of minimum temperature have less error than maximum temperature, but they are not be improved considerably after post-processing.