Application of fuzzy system and fuzzy clustering in climatology (temperature zoning of Chaharmahal&Bakhteyari province)

  Application of fuzzy system and fuzzy clustering in climatology (temperature zoning of Chaharmahal&Bakhteyari province)     D. Rahimi. ( * ),  Assistant professor of climatology, University of Isfahan, Isfahan, Iran.   email: dariush111353@yahoo.com     G. Valipoor.  M.A. student of climatology, University of Isfahan, Isfahan, Iran.     H. Yazdanpanah.   Assistant professor of climatology, University of Isfahan, Isfahan, Iran.   Received: 21 Desember 2009 / Accepted: 13 October 2010, 23-26 P     Extended Abstract   1- Introduction   Zoning techniques, including point and linear are spatial data analysis methods. Classification systems typically measure the distance and rely on probability theory should be . Classification is done using numerical techniques ( zero and one binary logic ) and non- numerical ( symbol processing field and fuzzy logic fuzzy set ). Temperature of the portion of solar radiation energy is absorbed to the surface effects and energy is converted to heat. This article has been tried using the system and fuzzy sets are determined temperature zone Chahar Mahal and Bakhtiari province .     Case study: A study area of Chahar Mahal and Bakhtiari is an area of 16,403 kilometers (Figure 1).The province with the average height of 2153 m and 3600 m elevation difference from sea level is considered a mountainous region that has a significant temperature differences .   Material: Data are used , including an annual average temperature data stations Province (Table ( 1). These data were measured and recorded in the period 1976-2005 . Table ( a) shows the average height and temperature stations province .     Table 1- station temperature profile   of e Chahar Mahal and Bakhtiari .   Station   Height   Average temperature   Average Max   Average Min   ImamGhaes   2400   10.3   18.7   2   Broujen   2197   9.1   18.8   2.8   Beheshtabad   1670   13.7   22.2   5.2   Dezak abad   2150   9.8   17   2   Solegan   2170   11.5   19.3   3.5   Shahrekord   2061   11.8   20.2   3.4   Avergan   2440   9.3   16.8   2.2   Dezak   2280   10.6   19.3   1.8   Kohrang   2285   8.35   19   2.6   Lordegan   1570   15.2   24.1   6.2   Marghmalek   860   20.1   27.4   12.7   Monj   1430   16.1   26.7   8.1   2-Methodology   Methods used in this article are based on fuzzy logic, multi-value method, nearest neighbor classification, indicators of class membership and the amount of members in the categories overlap. For nearest neighbor classification method has been used in the Euclidean distance fuzzy logic (Function 1):       Fuzzy Membership Function indicators are calculated using the following Function:     (2)   FBF z(x)=1 / { 1+(Z (X) – b1—d1/d1)}ifz(x)< (b1+d1)     (3)   FBFz(X) = IF(b1+d1) z(x) (b2-d2)     (4)   FBFZ(x) =1/{1+(Z(x) – b2 –d2/d2}Ifz(x)> (b2-d2)     Functions 2, 3 and 4 are marked degree of membership (FMF) Edit in clusters. With the help of these functions can be created between users and insert the terms of their membership in a state of optimal and possible.     3- Discussions   Fuzzy inference rules package (Rule Base) created is the most important step in the use of fuzzy logic. Therefore, based on certain principles of physics, was created using the correlation relationship between temperature and its effective factors such as altitude, relative humidity, latitude, longitude and precipitation depending on fuzzy inference rules. Set of fuzzy inference rules include: If the increased latitude then the temperature will be relatively low. If the increase of longitude then the temperature will be relatively low. If the increased of height then reduce temperature extreme. If the increases of relative humidity, then adjust the temperature.   Inference rules based on closed FBF index and temperature are three groups in the province of Chahar Mahal and Bakhtiari. Figure 1 shows the temperature groups based on FBF Other factor affecting the temperature is classified index ASW. This indicator shows the influence of different groups on each other. So the index value (ASW) in the group is equal 0.559, 0.0189, 0.5176 and the average index equal to 0.335 in the province. Figure (2) shows the temperature groups in Chahar Mahal and Bakhtiari.   .       Figure 1- The temperature groups based on FBF.       Figure 2- The temperature groups based on AWS.       4- Conclusion   The results show that in chahar mahal&bakhteari province three different temperature zones including: Shahrekord with average temperature 11-13 (c0), (cold zone), Koohrang with average temperature 8-11(c0) (very cold), and Lordegan with average temperature 14-20 (c0) (temperate and semi-warm).The first zone with ASW and area of 0.559 and 18.89%. Including Borojen and Pol-e-zamankhan,the second one with ASW and area of 0.018 and 53.97% including: Kohrang ,Imamgees,Solegan,Oregan and Dezak ,and the last zone with ASW and area of 0.517 and 27.13%inculoding Lordegan, Monj, Marghak, Beheshtabad and Barz.   Key words: Temperature, classification, fuzzy logic, fuzzy, fuzzy clustering, Chahar Mahal and Bakhtiari Province       References   Alborzi, Mahmoud. (1999). Introduction to Neural Networks, Second Edition, Sharif University of Technology.   Ghaffari, Ali. (1999). Fuzzy Thinking, Khaje Nasir Toosi University.   Ramesht, Mohammad Hussein. (1999). fuzzy geography and natural systems, Journal of Geographical Research, 52, and 53 195 Serial to 206.   Zahedi, Morteza. (1999). Fuzzy Set Theory and its application – Nashreh aneshgahi.   Zahydy Reza. (1380). uses fuzzy logic, neural networks - ISIRAN Institute.   Plan and Budget Organization. (1985). Master Plan Air and climatology section Chahar Mahal and Bakhtiari Province.   Meteorological Organization. (1952-2001). Statistical Yearbook of Chahar Mahal and Bakhtiari stations   Chahar Mahal and Bakhtiari Regional Water Company. (1966-2000). Planning and Studies Department Publishers.   Alijani, Bohlole. (1995). climate in Iran, Payam Noor University Publishers.   Alijani Bohlo and Kaviani, Mohammad Reza. (1992). Fundamentals of climate studies - Samt publication.   Ghayoor Hasanali and, Seyed Abolfazl Masoodian. (1997). Principles of Geographic Information Systems, University of Isfahan Publishers.   BjarneK.hansen &Denis Riordan. (2001). Weather Prediction Using Case based reasoning &Fuzzy Set Theory   BrianP.Mackey. (2004). Anon –liner fuzzy set technique for combing precipitation forcasts, 20thconferenceon Meteorological and Forecasting.   Burrough, P.A. (1989). Principles of geo graphical information systems for land resources assessment. Oxford University Press.   Burrough, P. A. (1989). Fuzzy mathematical methods for soil survey and land evaluation.Journal of soil science. No 40   Denis Riordan&BjarneK.Hansen. (2002). A Fuzzy Case_Based Model For Casting.