290 likes | 449 Views
Impacts of NAO on the Hydrology of Eastern Mediterranean . Ercan Kahya Istanbul Technical University. Outlines. A General View on NAO impacts in EM 2. Findings in M.E. Countries 3. Findings in Turkey 4. Conclusions. How do we expect NAO signals in Eastern Mediterranean (EM)? .
E N D
Impacts of NAO on the Hydrology of Eastern Mediterranean Ercan Kahya Istanbul Technical University Istanbul Technical University / Civil Engineering Department
Outlines A General View on NAO impacts in EM 2. Findings in M.E. Countries 3. Findings in Turkey 4. Conclusions Istanbul Technical University / Civil Engineering Department
How do we expect NAO signals in Eastern Mediterranean (EM)? Istanbul Technical University / Civil Engineering Department
NAO-Precipitation Variability in M.E. DJFM Composite Anomalies (1979-1998) (Cullen et al.: 2002) Most negative quartile of NAO index years Most positive quartile of NAO index years Istanbul Technical University / Civil Engineering Department
NAO-Streamflow in M.E. (Cullen et al.: 2002) Location of Streamflow stations Accumulated ppt for the period Nov-Apr / 1998–1999 to 2000–2001 • EOF analysis on 12 monthly subsamples of the interpolated records of 5 rivers • EOF 1 loadings at all stations are almost always positive, showing a common signal • EOF 1 accounts for 70% of total variance during DJFM and 60-90% during AMJJ • CONCLUSION: DJFM streamflow is most strongly influenced by the NAO Istanbul Technical University / Civil Engineering Department
NAO-Precipitation & Temperature in M.E. (Cullen & deMenocal: 2000) DATA: 211 temperature and 403 precipitation stations 30-year long Spatial correlation between the NAOSLP index and mean winter (DJFM) station (a) temperature; and (b) precipitation Istanbul Technical University / Civil Engineering Department
NAO-Precipitation in Mediterranean (Oleg Pokrovsky: 200?) Seasonal correlation of April-May (1995-2005) precipitation with February-March NAO (index leads by 2 months) Seasonal correlation of April-May (1965-1975) precipitation with February-March NAO (index leads by 2 months) Istanbul Technical University / Civil Engineering Department
NAO-Precipitation in Mediterranean (Oleg Pokrovsky: 200?) Seasonal correlation of June-July (1995-2005) precipitation with April-May NAO (index leads by 2 months) Seasonal correlation of Sep-Oct (1995-2005) precipitation with July-August NAO (index leads by 2 months) Istanbul Technical University / Civil Engineering Department
NAO Signals in Iran NAO-Regional Meteorological Drought • NAO negatively correlated with autumn rainfall [extreme autumn drought - the negative June–Aug NAO] • A spring drought is preceded by an Oct–Dec NAO > 0.5 • Winter droughts – not lag correlated with NAO • A wet autumn tends to occur when June–Aug NAO < −0.3 • Wet winter seasons – not lag correlated with NAO (Dezfuli, et al.: 2010) Istanbul Technical University / Civil Engineering Department
NAO Signals in Kuwait (Marcella & Eltahir: 2010) NAO-Rainfall • No significant correlation on a monthly or yearly time scale • Weak connections observed with both tropical and mid-latitude atmospheric systems • Reason: Kuwait is located in the transitional zone between the tropics and mid-latitudes Spatial Correlation: CRU’s Kuwait rainfall & surrounding rainfall for November–April Istanbul Technical University / Civil Engineering Department
NAO Signals in Oman NAO-Rainfall (Charabi & Al-Hatrushi: 2010) • Wet and dry spells - composite analysis - period 1984-2007 – reanalysis data used • Circulation features over the Caspian region responsible for winter rainfall variability • Dry spells – NAO (+) weak correlation (r=0.22) • Wet spells – negatively correlated with NAO (r=0.45) • Reason: NAO (-) phase → enhanced westerly flow (850-hpa up to 200-hpa), bringing more moisture from Atlantic Ocean across the Mediterranean area Annual average of rainfall 1984–2007 Winter average of rainfall 1984–2007 Istanbul Technical University / Civil Engineering Department
NAO Signals in Israel NAO-Rainfall Intensity (Yosef et al.: 2009) • Six rainfall categories - rainy season (Oct-May) - period 1950-2003 • A (light: 0-4 mm/d), B (moderate), A&B (light to moderate), C1 (moderate-heavy), C2 (heavy), and D1 (heavy-torrential: 64-128 mm/d) • NAO - significant impacts in center and south parts Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey (Cullen & deMenocal: 2009) NAO-Precipitation, Temperature &Tigris–Euphrates Streamflow Data period: 1930–1995 Stations: 23 temp & 27 ppt Correlation btw the NAOSLP index and Turkish winter temperature index (b) Turkish winter precipitation index (c) DJFMA average streamflow of the Euphrates (filled circles) and the Tigris Rivers (open circles) Results: 18% of the temperature variance is linearly related to theNAOSLP index 27% of the Turkish winter precipitation variance is linearly related to the NAOSLP index Istanbul Technical University/Civil Eng Dept.
NAO Signals in Turkey NAO-Tigris–Euphrates Streamflow (Cullen & deMenocal: 2009) • More profound NAO influence when NAO-extrema years are considered • At Keban station, the Euphrates streamflow exhibits 40% variability about • the 35-year mean value (663 m3/s) associated with NAO extrema Istanbul Technical University/Civil Eng Dept.
NAO Signals in Turkey NAO-Rainfall (Turkes: 1996) Rainfall in Turkey is characterized by two major wet periods: [1940 – 1948] and [1962 – 1970] by two major dry periods: [1971 – 1974] and [1989 – 1993] with severe and widespread droughts occurring in 1973, 1984, 1989 and 1990. • When compared with the NAOSLP index shown in figure (a) in page 12, aconnection between wet periods and a negative NAO, and drought periods and a positive NAO. NAO-Temperature (Turkes & Erlat: 2009) Investigated monthly mean winter (DJF) air temperature series at 70 stations in Turkey During NAO+ phase: - Northeasterly circulation increased, thus spatially coherent and significant cold signals During NAO- phase: - Spatially coherent and significant warm signals - Increased cyclonic activity and associated increased westerly and southwesterly circulation causing warm maritime air advection over the Mediterranean basin toward Turkey Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey (Karabörk, Kahya, Karaca: 2005) NAO-Precipitation NAO-Streamflow Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey (Karabörk, Kahya & Karaca: 2005) NAO-Tmax NAO-Tmin Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey NAO-Streamflow PC1 (Kalayci & Kahya: 2006) • Basic streamflow anomaly pattern • Reproduces 38.2% of the total variance • Period 1964-1994 Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey (Kalayci & Kahya: 2006) NAO-PC1 • PC1 as indicator of general level of streamflow • Highly significant relation for Annual and JFM Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey NAO-Lake Levels (Kucuk, Kahya et al.: 2008) • 7 lakes selected among 25 lakes with the period 1960-2002 • Using continuous wavelet transform analysis, - Relations Lake-level Periodicities & those of NAOI Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey NAO-Lake Levels (Kucuk, Kahya et al.: 2008) • Periodicities with 5- to 24-year scale level in Iznik, Beysehir & Egirdir lakes • Uluabat, Sapanca & Tuz lakes are in-phase relation with NAO periodicities less than 10-year & found to be more prone being affected by NAO Istanbul Technical University / Civil Engineering Department
(Sarlak & Kahya: 2006) NAO Signals in Turkey NAO-PDFs of Critical Droughts •Göksu River / Annual data (1936 -2000) • Two periods for NAO+ & NAO- (1936-1971)&(1972-2000) Model: A first-order Markov chain selected based on AIC & BIC Result: Expected critical drought durationsE(L): For example, say L=25, then E(L)=14.34 years for the NAO+ period, whereas E(L)= 6.17 years for the NAO- period. Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey NAO-PDFs of Critical Droughts (Sarlak & Kahya: 2006) Observed Transitional Probabilities for the Entire Period 1936–2000 • Results: • - Surplus probability P(s), wet flow conditions, increases to 0.583 for the NAO- period • The reverse is true for deficit probability • P(d) • Probability P(d/d), characteristic of negative anomaly followed by another negative anomaly, takes on a remarkably high value of 0.812 for the NAO- period Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey (Guney, Kahya et al.: 2008) NAO-Annual Cycle Modulation NAO Index Categorization Çoruh Basin (Northeastern Turkey) Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey NAO-Hydrologic Regionalization (Kahya et al.: 2010) Data Period: 1970-1997 Data Period: 1962-1971 NAO POSITIVE PHASE – 83 STATIONS NAO NEGATIVE PHASE– 40 STATIONS Method: Cluster Analysis Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey (Hydrologic Regionalization) NAO+ Phase NAO- Phase Qmax Qave Qmin
NAO Signals in Turkey Cyclone Track Climatology (Karaca et al.: 2000) DATA: Daily surface and 500-hPa charts for the period 1979-1994 PATH 1: Originates from the southwestern Russia PATH 2: Originates from the Balkans PATH 3: Originates from the Genoa Gulf PATH 3a: Moves to the northeast direction PATH 3b: Moves towards the east PATH 4: Originates from the western/middle Mediterranean or from southern Genoa Gulf or from north of the Sahara Desert Istanbul Technical University / Civil Engineering Department
NAO Signals in Turkey (Karaca et al.: 2000) Cyclone Track Climatology (a) Frequency of cyclones over Turkey (b) Cyclone numbers in months RESULT: For Turkey’s Mediterranean type of climate: the number of cyclones is the lowest in the summer and highest in the winter MISSING INDICATION: The period 1979-1994 corresponds the NAO+ period, so it is consistent to observe insignificant decreasing trend implication in the first panel Istanbul Technical University / Civil Engineering Department
Conclusions • NAO signals are detectable in Turkish climate/hydrological surface variables • No evidence for rainfall in Kuwait • Wet spells in Oman rainfall are significantly and negatively correlated with NAO • Significant rainfall intensity relations with NAO in central & southern Israel • Autumn rainfall/drought and spring drought in Iran significantly correlated with • NAO • Suggestion: A regional study in EM as a whole with dense data network • including major climate variables is needed … Ercan Kahya Istanbul Technical University Istanbul Technical University / Civil Engineering Department