See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/288079430
Regional trends in Turkish pan evaporation
Article in Journal of Food Agriculture and Environment · July 2012
CITATION
1
READS
52 3 authors, including:
Some of the authors of this publication are also working on these related projects:
Re-evaluation of trends in annual streamflows of Turkish rivers for the period 1968-2007View project
HORTICULTURAL RESEARCH INSTITUTE DIRECTORATE, Soil and Water Resources Tarsus Location, Dept. of Water ManagementView project Fatih Topaloğlu
Cukurova University 23PUBLICATIONS 186CITATIONS
SEE PROFILE
Mete Özfidaner
Republic of Turkey Ministry of Agriculture and Forestry 17PUBLICATIONS 33CITATIONS
SEE PROFILE
All content following this page was uploaded by Fatih Topaloğlu on 06 February 2017. The user has requested enhancement of the downloaded file.
960 Journal of Food, Agriculture & Environment, Vol.10 (3&4), July-October 2012 www.world-food.net Journal of Food, Agriculture & Environment Vol.10 (3&4): 960-962. 2012
WFL Publisher
Science and Technology
Meri-Rastilantie 3 B, FI-00980 Helsinki, Finland
e-mail: info@world-food.net
Regional trends in Turkish pan evaporation
Fatih Topaloğlu*, Mete Özfidaner and Funda Aydin
University of Çukurova, Faculty of Agriculture, Department of Agricultural Structures and Irrigation, 01330 Adana, Turkey *e-mail: topaloglu@cu.edu.tr
Abstract
Regional trend analyses of monthly pan evaporation data were conducted using the regional average Mann-Kendall test with both serial and cross- correlation at 66 sites in Turkey for 1975-2006. Regional increasing trends identified in monthly data have appeared to be predominantly insignificant all over Turkey with the strongest magnitude in May through August. However, majority of the regions exhibited insignificant decreasing trends in April. Increasing regional trends were in the Eastern Anatolia, Mediterranean Sea, Central Anatolia, Southeastern Anatolia, Marmara and Black Sea regions and decreasing trends in the Aegean region. When averaged the Sen slopes across the sites in a region over a 30 year period, the increase in monthly pan evaporation ranged between 2.7 and 44.1 mm or 2.1-25.7% whereas the decrease varied between 1.9 and 19.8 mm or 0.7-19.7% throughout Turkey. The implication of these results is that the terrestrial surface in Turkey has, on average, become more arid in the recent 30 years for May through October except the Aegean region.
Key words: Pan evaporation, Turkey, trend detection, regional average Mann-Kendall test. Received 10 June 2012, accepted 20 September 2012.
Introduction
Evaporation is one of the most important components of the hydrological cycle. It is also an important climate element mainly influencing the availability of water, particularly for agriculture and other fields. In Turkey, water is the most critical resource especially for agricultural use with a 74% share 1 and changes in water resources associated with climate change may have a critical impact on evaporation, hence water availability 2. Therefore, understanding of the possible impacts of climate change on evaporation is of great importance for Turkey where semi-arid climate prevails and water resources are not properly developed. Higher evaporation rate creates more arid environment while downward trend of evaporation results in a more humid environment. Pan evaporation is important because it gives useful clues to the direction of the change in actual evaporation 3 and is an estimate of potential evaporation since moisture is not a limiting factor in an evaporation pan 2. Furthermore, investigation of climate change effects on the variables of evapotranspiration can be effective in determining appropriate adaptation strategies for mitigating the probable damage from these effects 4. One reasonable way to understand whether or not evaporation has been impacted by climatic changes is to investigate if changes exist in historic evaporation records 5. However, recent studies of climate change in Turkey have focused primarily on precipitation, temperature and streamflow rather than evaporation. Besides, most trend detection studies ignored the role of temporal (serial) correlation and/or spatial (cross) correlation among datasets 6. Yue and Wang 5 developed a methodology namely regional average Mann-Kendall (RAMK) statistic that takes into account these affects applied by some other researchers later for streamflow 7 and precipitation 8, for only serial correlation affect in evapotranspiration data 9.
An analysis of Turkish pan evaporation trends taking serial
and cross-correlation among sites into consideration at the national scale has not previously been undertaken. The purpose of this study, therefore, is to analyse monthly pan evaporation data in seven homogeneous geographical regions of Turkey during 1975-2006 to identify changes through time using the regional average Mann-Kendall statistic with both serial and cross-correlation.
Materials and Methods
Turkey consists of seven geographical regions 10 having a drainage area of 780,576 km2 and is a large peninsula surrounded by the Mediterranean Sea in the south, the Aegean Sea in the west, and the Black Sea in the north (Fig. 1). Most of the precipitation falls during late autumn, winter and early spring due to main precipitation sources of moist air masses coming from the Atlantic Ocean and the Mediterranean Sea 11. For more details on the climatology of Turkey, readers are referred to Türkeş 11.
Pan evaporation data were available from the months of April to October for 1975-2006 at 66 meteorological sites (MS) located throughout Turkey. The data were observed and recorded by the Turkish State Meteorological Service which allows to use the data for only after 1975. Sites were selected based on the spatial distribution, availability of the longest possible records having common periods within the period after 1975 and completeness of the records. Fig. 1 shows the location of each meteorologic station and indicates the geographical region for each station. Stations are broadly distributed, not only across the country but also across the geographical regions of Turkey; the exceptions are the EA and BS regions. This is mainly due to a greater number of missing values and shorter record lengths in the regions. There are 7 MSs in the SEA, 7 MSs in the EA, 8 MSs in the MAR, 8 MSs in the BS,
Journal of Food, Agriculture & Environment, Vol.10 (3&4), July-October 2012 961
10 MSs in the MED, 12 MSs in the AEG and 14 MSs in the CA region. From the available 66 sites, 44, 64, 66, 64, 65, 64 and 58 stations were utilized for April through October, respectively. April includes 22 fewer sites located mostly in the CA and the EA regions. Therefore, sites that were not used for any month in the regional anaysis are also shown in Fig. 1.
Methodology: The rank-based Mann-Kendall non-parametric test has
been commonly used to assess the significance of trends in hydro- meteorological time series 5. The Mann-Kendall test is based on the null hypothesis that a sample of data is independent and identically distributed, which means that there is no trend or serial correlation among the data points. However, the presence of serial correlation and/or cross correlation in the data will affect the ability of statistical tests to assesss the significance of trends MK test 5. Therefore, Yue and Wang 5 proposed a regional average Mann-Kendall statistic that properly takes into account these effects in the assessment of the field significance of trends. Besides, lag-1 serial correlation coefficients 7 on the Mann-Kendall statistic for pan evaporation data were considered. In the case of trend, the serial correlation coefficient is affected and therefore a non-parametric robust estimate of the magnitude of the slope 12 is first eliminated from a time series and then the lag-1 serial correlation coefficient is recalculated. Further details on the regional
trend detection methodology can be found in Yue and Wang 5.
Results and Discussion
The assessment results by the RAMK test at the significance level of 5% are presented for the months of April through October in Table 1 in which the directions of pan evaporation trends are also provided regionally.
Results of the regional trend analysis revealed that only 24.5% (12 times over 49) of the monthly pan evaporation trend displayed an insignificant decreasing trend while 75.5% of stations were characterized with increasing trend. In addition, very few (5 times) of the increasing trends were found statistically significant at the significance level of 5%, whereas no significant decreasing trend in pan evaporation was observed. The regional trends in pan evaporation did not vary from regions to regions and from month to month with no immediately obvious spatial and temporal pattern except the AEG region. The implication of this result is that there is an increase of aridity across Turkey except the AEG region. Besides, the general increasing trend in pan evaporation identified in the study is consistent with the expectation that a generally warming climate would increase the evaporation from terrestrial open water bodies.
Furthermore, monthly results showed that the regional increasing trends have appeared to be predominantly insignificant all over Turkey with the strongest magnitude in May through August. The most number of regional insignificant increasing trends were identified for May and July (6 times over 7), August-October (5 times each), June (4 times) compared to the month of April (1 time). An upward regional trend was mainly identified for the EA and MED, the CA and SEA, MAR regions, partly for the BS region and the AEG region. On the contrary, most of the regions showed insignificant decreases
Months Geographical Regions
MAR AEG MED SEA CA BS EA
April -0.487 p -0.292 p 0.369 n -1.423 p -0.270 p -1.178 p 2.265*n May 1.204 n 0.865 n 1.837 n 0.714 n 1.658 n 0.889 n 2.241*n June 0.257 n 0.536 n 1.982*n 1.105 n 1.442 n -0.144 p 2.020*n July 0.914 n -0.138 p 1.343 n 0.399 n 1.444 n 0.733 n 1.589 n August 2.027*n -0.143 p 0.791 n 1.149 n 1.249 n 1.178 n 1.230 n September 0.183 n -1.533 p 0.291 n 0.291 n 0.140 n -0.728 p 0.532 n October -0.007 p -1.261 p 0.510 n 1.230 n 1.188 n 0.116 n 1.398 n
Table 1. The results of the RAMK test for pan evaporation in geographical regions of
Turkey.
Note: (*) represents that the trends are field significant at the 5% significance level. Arrows show pan evaporation direction: ↓, refers to downward trend and ↑, refers to upward trend.
Figure 1. Location of the sites for regional trend analysis in Turkey. Numbers in the brackets denote the number of month that
962 Journal of Food, Agriculture & Environment, Vol.10 (3&4), July-October 2012
References
1Pamuk Mengu, G., Akkuzu, E., Anac, S. and Sensoy, S. 2011. Impact of
climate change on irrigated agriculture. Fresenius Environmental Bulletin
20(3a):823-830.
2Burn, D. H. and Hesch, N. M. 2007. Trends in evaporation for the
Canadian Prairies. Journal of Hydrology 336:61-73.
3Ogolo, E. O. 2011. Regional trend analysis of pan evaporation in Nigeria
(1970 to 2000). Journal of Geography and Regional Planning
4(10):566-577.
4Shadmani, M., Marofi, S. and Roknian, M. 2012. Trend analysis in
reference evapotranpiration using Mann-Kendall and Spearman’s Rho tests in arid regions of Iran. Water Resources Management 26:211- 224.
5Yue, S. and Wang, C. Y. 2002. Regional streamflow trend detection with
consideration of both temporal and spatial correlation. International Journal of Climatology 22:933-946.
6Douglas, E. M., Vogel, R. M. and Kroll, C. N. 2000. Trends in floods
and low flows in the United States: Impact of spatial correlation. Journal of Hydrology 240:90-105.
7Topaloğlu, F. 2006. Regional trend detection of Turkish river flows.
Nordic Hydrology 37:165-182.
8Topaloğlu, F. and Özfidaner, M. Regional trends of precipitation in
Turkey. Fresenius Environmental Bulletin (in press).
9Li, Y., Horton, R., Ren, T. and Chen, C. 2010. Prediction of annual
reference evapotranspiration using climatic data. Agricultural Water Management 97:300-308.
10Tayanç, M., İm, U., Doğruel, M. and Karaca, M. 2009. Climate change
in Turkey for the last half century. Climatic Change 94:483-502.
11Türkeş, M. 2003. Spatial and temporal variations in precipitation and
aridity index series of Turkey. Mediterranean Climate: Variability and Trends. In Bolle, H.-J. (ed.). Regional Climate Studies. Springer Verlag, Heidelberg, pp. 181-213.
12Sen, P. K. 1968. Estimates of the regression coefficient based on
Kendall’s tau. J. Am. Stat. Assoc. 63:1379-1389.
13Eslamian, S., Khordadi, M. J. and Abedi-Koupai, J. 2011. Effects of
variations in climatic parameters on evapotranspiration in the arid and semi-arid regions. Global and Planetary Change 78:188-194.
in pan evaporation only for April. Besides, there were very few regions with no significant decreasing trends in September and October. On the other hand, the EA region was determined to be the sole region exhibiting significant increase from April to June followed by the MAR region for August and the MED region for June.
Over a 30-year period, we averaged the Sen slopes across the sites used in a region in order to give a rough indication in the concerned months. The regional average Sen slopes yielded an absolute increase in the regional average pan evaporation from May through October all over Turkey except the AEG and BS regions with an increase in pan evaporation for some months. Increase (percentage of change from the average) in pan evaporation for 30 years from May through October ranged between 2.7 and 28.8 mm (2.2-16.0%) in the MAR, 5.4 and 28.8 mm (2.1-11.0%) in the SEA, 3.6 and 24.2 mm (2.3-17.7%) in the CA regions. In these three regions, decrease in April was 4.5 mm (5.0%), 18.9 mm (17.5%) and 11.7 mm (19.7%), respectively. In addition, the MED and EA regions exhibited completely an increase in pan evaporation during the study period varying between 4.5 and 23.3 mm or 3.1-14.7%, 16.2 and 44.1 mm or 8.8-25.7%, respectively. Besides, in the BS region, the Sen slope yielded an increase in pan evaporation by approximately 3.7-14.9 mm or 3.7-8.9% for May, July, August and October while decreasing by 8.1-9.9 mm or 5.7- 10.6%. However, the AEG region is the only region resulting in a general decrease varying from 1.9 mm or 0.7% in July to 19.8 mm or 11.6% in September whereas an increase of about 11.1 mm or 7.0% and 12.6 mm or 6.0% was obtained in May and June, respectively.
Conclusions
The increase in pan evaporation, especially in May through August, tend to cause soils to dry more and can therefore have a negative impact on water resources. In other words, the results of this study displayed an increase in irrigation water requirement for agricultural use across Turkey due to evaporation increase. In case of water limitation, crop will not function with high productivity and also lead to economic loss due to the importance of agricultural activity in most areas of Turkey. It is necessary to use alternative irrigation techniques in agriculture such as deficit irrigation techniques and so on 13. Therefore, the findings are helpful for future water resources planning and sustainable agricultural development strategy not only at the local scale but also at the regional scale.
View publication stats View publication stats
