تحلیل خشکسالی کوتاه مدت در استان گلستان با استفاده از زنجیرۀ مارکف و روش SPI

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانشجوی دکتری اقلیم شناسی، دانشگاه آزاد نور ، نور، ایران

2 استادیار، گروه جغرافیا اقلیم شناسی دانشگاه آزاد اسلامی واحد نور، نور، ایران

3 دانشیار ، گروه جغرافیای طبیعی دانشگاه آزاد اسلامی، واحد رشت، رشت، ایران

4 دانشجوی دکتری اقلیم شناسی، واحد نور ،نور ، ا یران

چکیده

خشکسالی از موارد تغییر پذیر شرایط اقلیمی است که تعاریف مختلفی از آن ارائه شده در حالت کلی این پدیده حاصل کمبود بارش در یک دورۀ زمانی معمولاً ماهانه، فصلی و یا بیشتر است.امروزه با توجه به وقوع خشکسالی ها و آسیب های عمده به بخش های کشاورزی و منابع آبی نیاز در این است که ، برنامه ریزی مدونی برای کاهش اثرات آن ارائه نمود . هدف تحقیق حاضر تحلیل خشکسالی و بررسی احتمال وقوع آن در استان گلستان می باشد برای بررسی وضعیت خشکسالی از روش شاخص SPI و برایبررسی احتمال وقوع از روش زنجیرۀ مارکف مرتبۀ اول استفاده گردید. داده های مورد استفاد طی دورۀ آماری 30 ساله شامل 2 ایستگاه سینوپتیک2008-1979م و 8 ایستگاه باران سنجی استان طی دورۀ آماری 86 _57است شاخص‌های خشکی منطقه شامل شرایط خشکسالی‌ها، احتمالات ساکن و اقلیمی فصول خشک و تر ، طول دوره ها ترسالی و خشکسالی و احتمال خشکسالی های فصلی بررسی ومورد تجزیه وتحلیل قرار گرفته است
.نتایج حاصل نشان می‌دهدکه: طول دوره های خشک نسبت به مرطوب در کلیه ایستگاهها بیشتر بوده و از آغاز دهه 2000 م تدوام آن در ایستگاهها نسبت به سالهای قبل افزایش یافته است. خشکسالی های به وقوع پیوسته در اغلب ایستگاهها درفصولی ست که مرطو بترین فصل در کل دوره تحقیق بوده برای نمونه در ایستگاه گرگان شدیدترین خشکسالی در زمستان با شاخصSPI 98/2- اتفاق افتاده است.دوره های خشکسالی در کوتاه ترین زمان 5 فصل (1/1 سال) و دربیشترین زمان 21 فصل (5/1 سال) تدوام یافته است با توجه به ماتریس های احتمال انتقال در تمامی ایستگاه های مورد مطالعه به جزء تیل آباد احتمال وقوع فصل خشک بعد از فصل خشک بالای 50 درصد می باشد که بیشترین آن ایستگاه هاشم آباد گرگان با احتمال 66 درصد است که نشان دهنده دوام خشکسالی و احتمال بالای وقوع آن می باشد

کلیدواژه‌ها


عنوان مقاله [English]

The Analysis and Prediction of Short-term Drought by Using of Markov Chain and SPI Method in Golestan Province

نویسندگان [English]

  • Abdol hafez penahi 1
  • GholamrezaJanbaz Ghobadi 2
  • Parviz Rezaei 3
  • ali asgharzadeh 4
1 Ph.D. Student, Islamic Azad University, Nour, Branch
2 Assisitant Prof of Geography department, Islamic Azad University, Noor branch, Noor, Iran
3 Associate Prof. Islamic Azad University, Rasht , Iran
4 Ph.D. Student, Azad University, Noor Branch
چکیده [English]

Introduction:
Due to the growing trend of climate change in recent years and the occurrence of accident-causing events such as glaciation, storms, hail, drought, etc., the need for more analysis and forecasting of this type of climate-causing hazards is felt.
Since Golestan province is one of the agricultural hubs and there are many forests and pastures in it, managers, farmers, and ranchers must anticipate the occurrence of droughts to reduce possible losses during accidents and disasters. Therefore, in this research study, by selecting different stations and distributing the appropriate location in the selection of stations from the data of synoptic stations and rain gauges, evaporators using common annual data 1979-2008 AD and 56-57-85-86 to analyze and predict drought paid.
Materials and Methods:
The research place of Golestan province with geographical coordinates is 37.2898 ° N, 55.1376 ° E. In this research, in the first stage, 6 stations from meteorological stations of Golestan province (Gorgan, Gonbad, Bandar-e-Turkmen, Aq Qala, Kalaleh, Aq Togha) were selected with 36-year annual data. Precipitation data lacked independence and homogeneity of data. Thyer (2000) notes in his research that it takes at least 120 years to simulate the annual rainfall of an area using the Markov chain. Thyer et al. (2006)
To continue from Hashemabad Semioptic Station of Gorgan and Gonbad with statistical years (1979-2008) 30 years old and Evaporation and Rain Metering stations of Maravatpeh, Tangarah, Tamar, Til Abad, Araz Kuseh, Tarshkoli, Incheh Borun, and Ghaffarhaji with statistical years (2005-2006 to 1977-1978) 30 years selected and quarterly data collected with the calculation of 4 seasons and 30 years of data reaches a total of 120 cases.
Findings:
In these stations, in the whole studied period, there were 1200, 566 wet seasons, and 634 dry seasons, which indicates the increase of droughts.
According to the probability transfer matrix of the first order of the studied stations, the highest wetting limit is related to Hashemabad Gorgan station with a stability of 64% and the lowest limit is related to Tilabad station with a stability of 39%.
probability matrices such as hashem abad gorgan in winter are expected to have p00 means that the probability occrunce of a dry month After a dry month, 56% and the probability of occurrence of p11 means a probability of occurrence of a wet month after a wet month is 53%. P10 means a 47% probability of occurrence of a wet month after a dry month and p01 means a 46% probability of occurrence of a dry mouth after a wet month which is the lowest probability of occurrence.
Results:
1- According to the 30 years studied, the dry season occurred 634 times and the wet season 566 times, indicating an increase in drought periods and that its durability has also increased.
2- There has been a long period of drought in Gorgan station since 2005. There was only one wet season in winter 2006, i.e. during the last 14 seasons, it was only one wet season. In this period, the most severe drought with SPI 3- index (extremely dry) is unprecedented in the studied stations. Has occurred.
3- In Gonbad station, like Gorgan station, a drought trend has been observed since 2005. During the last 14 seasons, only 4 seasons have been wet and the highest level of drought occurred in this period has reached close to SPI 2- (extremely dry).
Results of transfer probability matrices
1- In all studied stations, P00 (probability of occurrence of a dry season after a dry season) is above 50% except (Til Abad), the highest of which is related to Hashem Abad station in Gorgan with a probability of 66%. This shows. The duration of droughts in the province is long.
2- In Maraveh Tappeh, Tangarah, Tamar, Tilabad, Arazkuseh, Tarshkoli, Incheh Borun, Ghaffarhaji, Gorgan, and Gonbad P01 stations (probability of occurrence of a dry season after a wet season) 42, 50, 47, 51, 44, 47, 41, 48, 34, 52, percent, most of which are related to Gonbad station and the least of which is related to Hashemabad station in Gorgan.
3- The percentage of drought in Gonbad station is higher than other stations. In Maraveh Tappeh, Tangarah, Tamar, Til Abad, Arazkuseh, Tarshkoli, Incheh Borun, Ghaffarhaji, Gorgan, and Gonbad stations, the probability of P10 occurrence (ie the probability of occurrence of a wet season after a dry season) is 55-51, 60-54.5, respectively. 55-54,61-5,52-47, 55-54, 52-45, 53-52, 51-36, 54-51, the highest percentage is related to Tilabad station with 61% and the lowest is related to Gorgan station with 36%.
4- The highest percentage of P11 (probability of occurrence of a wet season after a wet season) is related to Hashemabad station with 64% and the lowest is in Tilabad station with 39%
Conclusion:
Because winter has the wettest season and the highest frequency of wet season, but in some years dry season has occurred continuously in most stations in this season, for example in Gorgan station, the most severe continuous drought occurred in winter 2006 with an index of -98-98 It has been found that it is unprecedented during the study period. If winter is the wettest season and the highest frequency of the wet season in the whole province, it seems that this type of change is due to climate change, which needs further investigation.
According to the studies and the results of drought analysis in Golestan province, it shows that the length of dry periods has increased, which is in full accordance with the findings of Daneshmand and Mahmoudi (2017). And that the duration of droughts has increased since 2000 and the highest intensity of droughts has occurred in the years 2000-2008 and since 1979-80, so that since 2005 in Gorgan, Gonbad, and Ghaffar Haji stations is more intense and long.
Since the minimum drought period is more than one year and the maximum is more than 5 years, the droughts in the province are increasing and changing, and this can pose many risks to the agricultural sector and water resources.

کلیدواژه‌ها [English]

  • Golestan
  • Drought
  • Markof Chain
  • SPI
  • Seasonal Drought
  1.  

    1. Hanafi, Ali and Khosh Akhlagh, Faramarz and Soltani, Ali. 1391. Analysis of droughts in Tehran province using SPI index and its prediction based on Markov chain model, geography and stability of environment No. 3, Volume 2 1391-1002.
    2. Dostan, Reza .1395. An Analysis of Some Times of Iran in the Last Half Century, Climate Studies 1394, No. 23 18-1.
    3. Siasar, Hadi and Farideh Shahrdarazi, 2015, Drought Monitoring and Prediction Using SPI Drought Index and Markov Chain (Case Study: Iranshahr County), 3rd International Conference on New Achievements in Engineering and Basic Sciences, Europe, Ukraine, Earth Research Center Cow, London Society of Engineering.
    4. Sadeghia Nia, Alireza ,hjazizadeh, Zahra, Hamidianpour, Mohsen and Poursarbandan, rahebeh. 1392. Estimation of drought and drought probabilities using SPI index and Markov chain model. Case study Tehran, Geographical Space No. 43, 1392 81-65.
    5. Mehri, Sonia , Haji, Khadijeh , Alizadeh, Varia, Mostafazadeh, Rauf .1394. Evaluation of spatial variations of intensity of meteorological drought periods in different time scales in Kurdistan province, Geographical Information (Sepehr) Volume 26 no 102 1396 162-151.
    6. Moghadam, H and Partners. 1380. Drought monitoring based on the profile of decay and normal SPI in Sistan and Baluchestan province, Volume 3 80-69 Proceedings of the first conference of Zabul water crisis response strategies.
    7. Yousefi, N and Hejam, S and Irannejhad, P., 1384. Estimation of drought and probability probabilities using Markov chain and normal distribution (Qazvin case study) Geographic researches No. 60, Summer 86 28 -121.
    8. Daneshmand, H. and P. Mahmoudi. 2017. Estimaion and assessment of temporal saatability of erio dicities of droughs in iran.water resources management, volume 31issue11, sepember pp3413-3426.
    9. Iqtedari, Mehrnoosh, Bazrafshan, Javad, Shafiei, Maryam, Hijabi, Somayeh. Prediction of Streamflow Drought Using SPI and Markov Chain in Kharkheh’s Basin. Journal of Soil and Water Conservation Research 1395; 23(2): 115-130. doi: 10.22069/jwfst.2016.3058.
    10. Kamasi, Mehdi, Malek Mahmoudi, Mehdi, Montaseri, Hossein.1396. Drought forecasting by SPI and EDI indices using ANFIS method based on C-mean and SC clustering (Case study: Kohgiluyeh and Boyer Ahmad Province) Agricultural Meteorology,: 1396; 5(1): 36-47. doi: 10.22125/agmj.2017.54982.
    11. Madanchi, Peyman, Shahedi, Kaka, Habibnejad, Mahmoud, Soleimani, Karim, Fatehi Marj, Ahmad. Zoning Climatic Drought and Drought Magnitude Using SPI Index and Kirijing Geostatistical Method (Case Study: Kerman Province). Iranian Journal of Irrigation and Water Engineering, 1398; 10(2): 205-228. doi: 10.22125/iwe.2019.100753
    12. Eullmoon, BoomRyoo, S, Gilkwon. 1994, Markov chin model for daily perepition accurrence in siuth korea, international jornal of clima tology 14, 1009 – 1016
    13. Thyer, M. A, P. H. D thesis, university of Newcastle, Australia (2000, modeling lonj-termpesistncc in hydrolojical tim series.

     

    1. Thyer, M. A, A. Y. Forst and G. akuczera (2006) parameterestimation and modelidentificotion for stochastic models of annuel hydrolojical data: is the observed record, lonj enoujh. Journal of hydrology. volnes 330. issues 1-2, 313-327.

    T. skiris, G. vanjelis.H (2004) towared sadrought watch system based on spatial SPI. water resources manajment, 18:1.