LEE- Geomatik Mühendisliği Lisansüstü Programı

Bu topluluk için Kalıcı Uri

http://hdl.handle.net/11527/19289

Gözat

Konu "Akdeniz Bölgesi" ile LEE- Geomatik Mühendisliği Lisansüstü Programı'a göz atma
  • Öge
    Evaluation of grid based precipitation products over the Mediterranean region in Turkey
    (Graduate School, 2022-02-08) Hişam, Enes ; Şeker, Dursun Zafer ; Danandeh Mehr, Ali ; 501191658 ; Geomatics Engineering
    Precipitation is an important part of the hydrological and energy cycle, as well as a key input for many applications in hydrology, climatology, meteorology, weather forecasting, and socioeconomics. It is also a crucial factor to consider when evaluating the consequences of climate change at different spatial scales. As a result, precise precipitation estimation is critical for all the aforementioned applications. However, such accurate estimation is difficult due to the variability of precipitation throughout time and space. The main sources of precipitation data include rain gauge stations and weather radar stations as ground-based observations, as well as grid-based precipitation products. However, ground-based rainfall estimates lack spatial coverage, which is a significant issue for regional and global applications. In recent decades, considerable efforts have been undertaken to generate gridded precipitation products, resulting in a rise in the number of precipitation datasets at various spatial and temporal resolutions on a global or quasi-global scale. The key benefits of these products over ground-based stations (rain gauges and weather radars) are that they produce worldwide precipitation data with continuous and high spatial and temporal resolution, as well as public access to this data. Such information is especially important in developing countries or rural areas where weather radar or rain gauge data is scarce or of low quality. All gridded precipitation products, however, are subject to a number of errors, including sensor and algorithm faults. Furthermore, the precision of these products varies by place, season, climate, topography, and clouds. As a result, before applying gridded precipitation products in a specific location, a thorough and comprehensive assessment is necessary. In this study, an evaluation of six gridded precipitation products was performed over the Mediterranean region in Turkey from 2017 to 2021. The evaluation was performed at multiple temporal (monthly and annual) and spatial (grid and regional) scales. The precipitation data from the ground-based station (points) was upscaled to grids using Interpolation Weighting Average (IDW) and areal average techniques for the evaluation of the products at grid scale. To assess the gridded precipitation accuracy, 193 ground-based meteorolgical stations distributed through the study area were used. The products include (1) Integrated Multi-satellitE Retrievals for GPM (IMERG) with 0.1° spatial resolution (2) Group InfraRed Precipitation with Station data (CHIRPS) with 0.05° spatial resolution (3) Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks Cloud Classification System (PERSIANN CCS) with 0.04° spatial resolution (4) Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks - Dynamic Infrared Rain Rate near real-time (PDIR-Now) with 0.04° spatial resolution (5) PERSIANN- Climate Data Record (PERSIANN CDR) with 0.25° spatial resolution (6) ECMWF Reanalysis v5 (ERA5) with 0.25° spatial resolution. The results showed that at all spatial and temporal scales, CHIRPS, PERSIANN CDR, ERA5, and IMERG perform close to each other and better than PERSIANN CCS and PDIR. At all spatial and temporal scales, all products showed overestimation bias for low precipitation events whereas underestimation bias for heavy rainfall events. Except for PDIR, which showed an overestimation trend at both a monthly regional and annual grid scale for high precipitation occurrences. However, when compared to severe rainfall events, the performance of all products is better at low and moderate precipitation events. Therefore, the performance of the products was better in summer and spring months (March to October) than those of winter (December to February). At all spatial and temporal scales, CHIRPS, PERSIANN CDR, ERA5, and IMERG all perform similarly to each other and better than PERSIANN CCS and PDIR. So, after using some bias correction techniques, all four products could be promising to use as a complement to rain gauge stations for hydrological and environmental purposes in Turkey's Mediterranean region. However, taking into account the uncertainties of their data, PERSIANN CCS and PDIR could be employed for applications that require real-time precipitation data.