چكيده فارسي :
هدف از نگارش اين مقاله، بررسي تداوم روزهاي توفاني شهر زاهدان بر پايه زنجيره ماركف است. به منظور شناسايي روزهاي توفاني، داده هاي روزانه ديد افقي و سرعت باد ايستگاه هواشناسي زاهدان در دوره آماري (1390 - 1362) فراهم شد. سپس روزهايي كه سرعت باد از آستانه 15 متر در ثانيه فراتر رفته و همچنين ميدان ديد به كمتر از 1000 متر كاهش يافته بود به عنوان روز توفاني استخراج شدند. بر پايه مدل ماركف، ماتريس هاي فراواني روزهاي توفاني ديدباني شده و ماتريس‌هاي احتمال انتقال ماهانه، فصلي و سالانه حساب شد. سپس وابستگي روزهاي توفاني و غير توفاني به يكديگر، به همراه ايستايي و همگني آن‌ها آزمون شد و فراواني روزهاي توفاني مورد انتظار، طول دوره روزهاي توفاني و غير توفاني و توالي دوره هاي توفاني (n) روزه محاسبه شد. داده‌هاي ديده‌باني شده نشان داد كه سال 1386 خورشيدي با 78 روز توفاني و دو سال 1379 و 1372 با 4 روز توفاني، به ترتيب بيشترين و كمترين روزهاي توفاني دوره آماري را داشتند. در بررسي فصلي فراواني روزهاي توفاني مشخص شد كه بيشينه روزهاي توفاني در زمستان (306 روز) و كمينه آن در پاييز (57 روز) رخ داده است. اسفند نيز با 123 روز توفاني و آبان با 8 روز به ترتيب بيشترين و كمترين روزهاي توفاني را داشتند. بررسي دوره هاي توفاني يك تا هفت روزه نشان داد كه بيشينه تداوم دوره هاي توفاني در اواخر زمستان و اوايل بهار و كمينه آن در اواسط پاييز رخ مي‌دهد. بيشينه سرعت باد در زاهدان در سال 1381 با 40 متر بر ثانيه در فصل بهار رسيده است. ميانگين روزهاي توفاني ۲۴/۴ و با انحراف معيار ۱۹/۳ برآورد شده است
چكيده لاتين :
Natural hazards that occur consistently across the universe put at risk human lives and
properties. Extreme climate phenomena currently focused the attention of researchers, for
the risk of increased frequency; duration and increased sensitivity of climatic thresholds
have been increased. Storms, one of the most hazardous natural disasters, are considered in
the group of weather hazards. Storm is the most important parameter to understanding
climatic phenomena, which occurs by severe atmospheric turbulence and severe weather
disruption is a very severe flow of weather that cause falling trees, collapsing buildings and
breaking glass etc. and smashes vast areas. Storm, as one of the climatic phenomena plays a
main role in human lives and his/her present and future plans and is also one of the study
bases of environmental planning. The storms cause financial losses to facilities, networks
and residential areas and also the death and injuring people.
The purpose of this study is to analyze the continuity of storm days based on Markov chain
in Zahedan city. Zahedan is the capital of the province of Sistan and Baluchestan that located in South-East of Iran. This region has been dry climate and storm days are one of the most
important climatic characteristics. It is called a storm day if the maximum wind speed
exceeds 15 meters per second and horizontal visibility is less than 1000 meters.
Matherials and Methods
In order to identify the storm days or windy days, horizontal visibility and daily wind speed
data during 1983 - 2011 were obtained from meteorological station with 1369.9 m height
above the sea level in Zahedan city. First, accuracy survey and homogeneity test of daily data
of wind speed using Run Test was conducted. Based on Markov Chain model, the matrices of
the frequency of windy days observed were constructed and the matrices of the
probabilities of transition for months and seasons were calculated. In the next step, the
correlation of windy days on each to other, with the durability and homogeneity, was tested.
Moreover, the expected frequency of windy days, the period of windy and non-windy days,
and the sequence of the windy n-days for each month were calculated.
Discussion and Conclusion
A windy day forecasting method based on the use of discrete time Markov chain models is
developed starting from real wind speed time series data. It allows to directly obtaining in an
easy way an estimate of the wind speed distributions on a very short-term horizon, without
requiring restrictive assumption son wind speed probability distribution. Markov Chain
Model is analytically described. Finally, the application of the proposed method is illustrated
with reference to a set of real data. The observed data showed that 2007 had the highest
frequency windy days (78 windy days) and, in contrast, the years 2000 and 1993 had the
lowest windy days (4 days). In the seasonal study of the frequency of windy days, it was
found out that the maximum windy days occurred in winter (306 days) and the minimum in
autumn (57 days). Moreover, March with 123 and November with 8 windy days had the
highest and the lowest windy days. Conclusion
According to Markov Chain model, analysis of windy periods of one to seven windy days
showed that the maximum Persistence of windy days, occurred in the late winter and early
spring and the minimum was in mid-fall. The maximum wind speed occurred in the spring of
2002 (40 meters per second). The average of the windy days was 24.4 days, with a standard
deviation of 19.3 days.
Calculating the probability of storm days in the higher-order Markov chain shows that the
percentage of likely sequences of one to seven days with a big difference in the months of
the year. Storm duration in each period is inversely related to the frequency of storm days.
This means that increasing the percentage of smaller sequences, the probability of a larger
sequence is reduced.
Comparing the transition probability and climatic probability shows that the occurrences of
storm days have very little difference between conditional probability with Non-conditional
probability which is calculated using Markov Chain Model. Therefore it can be concluded
that the application of this model to determine the probability of storm days is sufficiently
accurate.
