كليدواژه :
خودهمبستگي فضايي , آماره موران و Gi , تحليل داده هاي محيطي , عوامل جغرافيايي , ناحيه خزري
چكيده فارسي :
از مهم‌ترين اهداف آمار فضايي، بررسي روابط مكاني داده هاي محيطي (بارش شديد) براي تحليل الگوها و وابستگي هاي فضايي است. در اين راستا تحليل اكتشافي داده‌هاي فضايي روش هايي را براي تمايز بين الگوهاي تصادفي و غير تصادفي فراهم مي آورد. در اين پژوهش با كاربرد آماره فضايي به تبيين الگوهاي مكاني بارش هاي شديد كه با پيامد هاي محيطي – انساني همراه است، پرداخته شد. در اين راستا مجموع فراواني ماهانه بارش‌هاي شديد فصول سرد ناحيه خزري در دو گروه آستانه صدك 95-90 و 99-95، از 385 ايستگاه اقليمي طي 2016- 1966 با كاربرد آماره موران و آماره Gi*به‌منظور تحليل خودهمبستگي فضايي استفاده شد. يافته هاي آماره موران كلي نشان داد كه رفتار حاكم بر بارش شديد داراي ساختار فضايي و به شكل خوشه‌اي است. بررسي هاي ماهانه فصول نشان داد كه به لحاظ فراواني وقوع بارش هاي حاصل از اين دو گروه بارشي، فراواني رخداد بارش هاي صدك آستانه 95-90 بيشتر از آستانه صدك 99-95 مي باشد. براساس نقشه‌هاي موران محلي نواحي با خود همبستگي‌هاي مثبت در هر دو گروه بارشي بيشتر در نواحي غربي، مركزي و الگوهاي حاصل از خودهمبستگي منفي در بخش شرقي قرار دارد. در بررسي روابط مكاني، آماره دو متغيره موران بين بارش شديد و عوامل جغرافيايي نشان داد كه تاثير عوامل جغرافيايي در فراواني رخداد فرين‌ها ضعيف‌تر از عملكرد و نفوذ سامانه‌هاي همديد است. درنهايت، در اين ناحيه كه اين بارش ها يكي از مخاطرات طبيعي است، شناسايي اين الگوها مي تواند در مديريت و برنامه‌ريزي و كاهش آسيب‌پذيري و افزايش سازگاري موثر باشد.
چكيده لاتين :
Introduction
One of the most important goals of spatial statistics is the study of spatial relationships of environmental data for the analysis of patterns and spatial dependencies. From behavioral aspects of precipitation as environmental data are extreme amounts of precipitation. Extreme event is the occurrence of a value of a weather or climate variable above (or below) a threshold value near the upper (or lower) ends of the range of observed values of the variable, that are associated with negative and environmental-human consequences (such as flood, drought, landslide, soil erosion, and physical damage to infrastructure (roads-dams), impact on human activities (settlement-agriculture-industry-services). In this regard, exploratory analysis of spatial data provides methods for distinguishing between random and non-random patterns. Precipitation extremes follow a geographical pattern like all other climate elements. Recognition of such patterns, specifically in those areas where people’s lives depend on precipitations can determine the success in environmental management as well as certainty in resources planning. The geographical position of Iran’s coastal region of Caspian Sea,(Adjacent to the Caspian sea, and the Alborz mountains), and the adjacent various geographic units (sea, plain and mountains),as well as their interactions with each other, provide appropriate conditions for extreme precipitation occurrences in this area. As researchers believe that the extreme precipitations in Caspian region is part of the intrinsic properties (Mofidi,2008). Regarding extreme precipitation in Iran’s coastal region of Caspian Sea, especially in eastern parts, is one of the natural hazards.the recognition of spatial auto-correlation of such a phenomenon can facilitate environmental planning and the reduction of vulnerability and also increasing adaptability capacities with such a disaster.
Materials and Methods
In order to analyze the auto-correlation of the sum frequency of monthly heavy precipitations of the study area, the 90-95, 95-99 percentile of precipitation for each pixel of the map is considered in both groups for cool seasons. Accordingly, 385 stations (synoptic, climatology, and rain gauge of Islamic Republic Organization of Meteorology, and rain gauge of the Ministry of Power) were studied during the time period covering 1966 to 2016. At first, the frequency of monthly extreme precipitation was plotted in the Surfer software. Then, spatial statistics techniques (global Moran index (function 1), local Moran (function 2), and Gettis-ord-J index (function 3) were used to analyze spatial auto-correlation features.
Function (1)
Function (2)
Function (3)
In the last step, the relationship between the spatial factors with the extreme precipitation frequency for each month was calculated using general Moran multivariate statistics (function 4).
Function(4)
Results and Discussion
One of the optimal methods for identifying the spatial distribution of climate events is to analyze the spatial relationships. The study aimed at determining the spatial pattern of the sum monthly precipitation frequency patterns in two groups (90-95, 95-99 percentile), using the spatial statistics techniques (global Moran index, local Moran, and Gi* index). The findings indicated that the global Moran index is statistically at 99% significant level. Results of the present showed that the dominant behavior in sum frequency of monthly extreme precipitation of the study area followed a cluster pattern in each group. The areas with positive auto-correlated clusters were in western and central regions and negative auto-correlated clusters were in the eastern parts. Gi* test approved the frequency of clusters with high and low values. The two-variable Moran's statistics between extreme precipitation and geographic factors showed that the effect of geographic factors on the spatial frequency pattern of extreme precipitation occurrence is weaker than the performance and influence of synoptic systems.
Conclusion
In general, it can be said that the Caspian region is more affected by the precipitation of the first group (90-95 percentile), which covers a vast area of the region. Results of the study showed that the dominant behavior in sum frequency of monthly extreme precipitation of study area followed a cluster pattern. The areas with positive auto-correlated clusters wاere in western and central regions and negative auto-correlated clusters were in the eastern parts. The comparison of the findings with those of previous studies showed that the geographical location of the study area, features of Alborz mountains, and also synoptic systems have affected the spatial auto-correlation frequency pattern of extreme precipitation occurrence. Finally, considering that heavy precipitation in the Caspian Sea region causes one of the natural hazards (flood), especially in the eastern parts with heavily-populated regions, recognizing the spatial patterns of this phenomenon can be very useful for planning environmental hazards and reducing vulnerability and increasing the compatibility
