reconstruction of maximum temperature variations in kermanshah province using tree rings

introduction iran is located in a semi arid region of the world. in recent years, dramatic changes in the form of increasing in maximum temperature and severe droughts have damaged water resources, local forests and agriculture. in the mid- and high latitudes of the northern hemisphere the warmest years have been recorded in recent decades (ipcc, 2007). in order to fully understand the climatic fluctuation in each region and its effect on regional ecosystems, long time monitoring of climatic parameters is necessary. using these gathered data and different methods and modeling, the regional climate change can be studied. establishment of meteorological stations in iran dates back not longer than 1950s. to face this challenge, indirect evidence of past climatic situation which is recorded in some living and non living structures can be used. tree rings are annually resolved natural archives that provide proxy data for palaeo-environmental studies and reconstructions of various climate elements. despite of long-lived tree species in the north and the west of iran, unfortunately dendroclimatology studies are still scare in the region. the zagros mountain range in the west of iran contain the largest oak woodlands in the country and these trees are the most abundant and important species in the zagros forests. in addition, one of the most important sites of quercus infectoria olive is located in kermanshah province. the relatively high longevity of these trees and their wide distribution make them potentially suitable for dendroclimatological approaches. our aim is to reconstruct variability of maximum temperature in kermanshah province based on oak tree rings. materials and methods faryadras site is one of the largest sites of quercus infectoria olive in kermanshah province, western iran. the region is a part of the central zagros mountains. in the region, precipitation usually falls during an eight month period from october to may whereas there is no effective precipitation in other months during a year. we extracted 20 cores from 10 trees at breast height using an increment borer. only the largest diameter trees with no obvious injury or disease were sampled. the samples were mounted on sample holders, coded, air dried and measured from bark to pith with a lintab5 measuring system at a resolution of 0.01 mm using the software package tsap-win. all growth curves were cross-dated by visual and statistical tests (sign-test and t-test) using the software package tsap-win. the raw ring-width series were standardized to remove biological growth trends as well as other low-frequency variations due to stand dynamics. we decided to use residual chronology (rc) were constructed with the arstan program. the reliability of the chronology was evaluated by the expressed population signal (eps; wigley et al., 1984). meteorological data for our study were available from the climate stations of kermanshah (33°21’n, 47°09’e; 1318 m a.s.l.). we used monthly and seasonal maximum temperatures from pervious january to current september of the meteorological station to calibrate the ring-width_ climate relationships during the common period 1951–2010. to assess the temporal stability of the climate calibration model, the instrumental data (1951–2010) were split in half (1951–1980, 1981–2010) for calibration and verification tests. due to high correlations between the site chronology and instrumental data, a linear regression model was used to reconstruct past maximum temperature variations. result and discussion statistical tests (sign-test and t-test) showed that all of the growth curves have the same trend in the faryadras site. the length of the chronology is 305 years (1705 – 2010), the first year with eps >0.85 was found in a.d. 1840. quercus infectoria tree rings show the high sensitivity to maximum temperature and maximum temperature is one of the limiting factors of the tree growths in faryadras site. maximum temperature has a negative effect on tree growths from previous january to current september. significant negative correlations between maximum temperature and the residual chronology were found in february, march, may, june, july (p<0.001) and august (p<0.005). negative effects of maximum temperature on quercus infectoria tree rings during the per-growing and growing season have also been noted by najafi et al., (2010). it should be noted that the negative effect of maximum temperature on tree growths in growing season is severe that the pre-growing season. increasing maximum temperature reduces water availability of trees due to enhanced evaporation. our results showed that cold periods accrued in 1842, 1848, 1858, 1864, 1874, 1876, 1885, 1890, 1940, 1950, 1957, 1973, 1982, and 1992 years. severe warm years prevailed in 1847, 1871, 1944, 1948, 1960, 1984, and 2009. conclusion a tree-ring width chronology was developed in kermanshah province. maximum temperature has a negative effect on tree growths in faryadras site during the pre-growing and the growing season. the greatest potential for the development of a climate reconstruction for kermanshah province was found for may–august maximum temperature of the growing season. due to the correlation coefficient between instrumental data and the site chronology, may–august mean maximum temperature of the growth year was used for reconstruction over the last 170 years (1840-2010). our reconstruction contains interannual to multi-decadal variability. further sampling and applying other reconstruction methods like stable isotopes or wood anatomical tree-ring parameters may allow the preservation of long-term climatic trends and the reconstruction of additional climate parameters in this region.