Presentation is loading. Please wait.

Presentation is loading. Please wait.

PRECIPITATION TENDENCIES & DIAGNOSTIC PRECURSORS IN THE UPPER DANUBE CATCHMENT PARALLEL TO GLOBAL WARMING (1974-2003) János Mika 1, Vera Schlanger 1,

Similar presentations


Presentation on theme: "PRECIPITATION TENDENCIES & DIAGNOSTIC PRECURSORS IN THE UPPER DANUBE CATCHMENT PARALLEL TO GLOBAL WARMING (1974-2003) János Mika 1, Vera Schlanger 1,"— Presentation transcript:

1

2 PRECIPITATION TENDENCIES & DIAGNOSTIC PRECURSORS IN THE UPPER DANUBE CATCHMENT PARALLEL TO GLOBAL WARMING (1974-2003) János Mika 1, Vera Schlanger 1, Blanka Bartók 2 & Gábor Bálint 3 1 Hungarian Meteorological Service, Budapest, 2 Babes Bolyai University, Cluj, Romania, 3 Water Resources Research Centre, VITUKI, Budapest with special thanks to Judit Bartholy & Rita Pongrácz, Eötvös Loránd University, Budapest; Emőke Borsos & Gábor Pándi, Babes Bolyai University, Cluj, Romania

3

4

5 Climate Change 2001:The Scientific Basis WHY NOT SIMPLY THE OAGCM OUTPUTS ?

6 REGIONAL SCENARIO APPROACHES 1*. Raw GCM outputs (interpolation) 2*. Empirical analogues or simple statistics similarity hypothesis: regional response depend on the measure of global warming, but not on its causes 3 +. Physical downscaling with embedded mezoscale models. 4 ++. Statistical downscaling, based on circulation patterns *tacled in the presentation + +additional remark to the + just one figure from the literature one in December 2003

7 ...... 50 N 45 N 10 E 20 E 30 E 50 N 45 N TsB... TsB - Transsylvanean Basin 76 st. precipitation GCM (ScenGen) Pressure gridpoints Cloudiness & OLR Sectors of elaboration (as of April 7, 2004)

8 Figures of 3-component Fourier approximation (mean of 76 individual statistics) 25 years mean C1 C2C1+C2 C3 C1+C2+C3 Mean fit71%19%91%3%94% Best fit95%58%100%47% 100% Worst fit17%1%32%0%74% Year by year C1 C2C1+C2 C3 C1+C2+C3 Mean fit33%18%51%15%66% Best fit93%83%98%87%99% Worst fit0%0%1%0%6% P(t i ) = Ao + + C1 + C2 + C3 + err. i = 1, 2, …, 12

9 Relation of Fourier-components

10 Regression from short series Regression from short series Method of instrumental variables, first applied by Groisman (e.g. Vinnikov, 1986) in climatology Y(t)=Y o +Y 1 (t) not 0 correlation with (t), Z(t) can be selected for instrumental variable, if it exhibits not 0 correlation with (t), correlation with observation errors of (t) and correlation with residuals of Y (t)  0 correlation with observation errors of (t) and  0 correlation with residuals of Y (t) if Z(t) exists, than cov (Y, Z) Y1  cov (,Z) Z:= t 76 stations available

11 ANNUAL MEAN CHANGES Decrease 0 Increase

12 Intra-region similarity of regression and their variation with the altitude Macro- Coeffi-Annual Winter Summer- region cient totalhalf-yr.half-yr. AlpsCorrel.0.5840.6670.524 C, D,Altitude--------- JLatitude---8.6--- 17 st.Longitude-5.5-8.64.8 W-Carp. Correl.0.3970.6770.356 A, I,Altitude---1.1--- H, GLatitude------5.3 29 st.Longitude-2.5-4.7--- E-Carp. Correl.0.6580.4080.670 -1.7-2.0-1.7B, E,Altitude-1.7-2.0-1.7 F, KLatitude--------- 30 st.Longitude-2.8----3.3 Fig. 1 Proportion of the minority of stations exhibiting different sign of precipitation changes within the 9 regions (columns) than their dominant fraction, as compared to the same proportion without grouping (background curves).

13 Winter half-year +4 0 +3 0 +20+1 0 0 -1 0 -2 0 -3 0 -4 0 -5 0 -6 0 Summer half-year

14 Example: Change field for Europe 0  20  40  E 60  S 50  40  Analysis of Regional Climate Change Scenarios for Hungary MAGICC/SCENGEN Wigley et al., 2001; run & elaboration: Schlanger et al., 2003

15 CHANGES FROM SCENGEN (7 - 16 models: inter-quartals) Középhőmérséklet [°C] Napi hőingás [°C]Max. hőmérséklet [°C]Min. hőmérséklet [°C] Felhőborítottság [%]Csapadékmennyiség [%]Légnyomás [hPa]Szélsebesség [%] 2050 2100 Mean temperature K KKK % % hPa Diurnal temp. range Max. temperatureMin. temperature Cloud coveragePrecipitation amountVapour pressureWind speed

16 Data for cloudiness Ground-based visual observations 1973-1996 (24 years) 172 observation srtations EECRA data-base (Hahn & Warren, 1999) GCM-output fields (clouds, etc. ) 7-16 GCM modell adatai (CCC-EQ, CSIRO1, CSIRO2, ECHAM3, HAD- CM2, UKTR, UKHI-EQ for clouds) 2050 - B1 scenario (1,1 K warming) (MAGICC/SCENGEN diagnostics) 1961-1990 reference period (0 change) Output fields at 5 X 5 deg. rectangles

17 Data for cloudiness (cont.) Outgoing longwave radiation (OLR) 1979-2000 (22 years) 2,5 X 2,5 deg. resolution NOAA/NCEP TIROS-N quasi-polar satellites, AVHRR sensors Main regulator of OLR: the cloud coverage - Correlation coefficients (N = 18 : 1979-1996) Orbital altitude:~ 850km Original resolution: 1 - 3 km Frequency of images: ~ 6 hours NOAA NOAA - TIROS

18 CHANGE OF CLOUDINESS IN EASTERN-EUROPE1973-1996 r = 0.749 d /dt = 0.021 K/yr 1979-2000 r = 0.736 d /dt = 0.022 K/yr

19 ...... 50 N 45 N 20 E 10 E 30 E Data: CRU - Norwich University,

20 VALIDATION OF PRECIPITATION TENDENCIES FOR 1999-2003 VALIDATION OF PRECIPITATION TENDENCIES FOR 1999-2003 Transsylvanean Basin

21 Absolute extremes (high & low): Transsylvanean Basin Precipitation extremes (1/yr) Runoff extremes (1/yr) 9 stations (108 extremes each side), 19 stations (228 extremes each side), between 1974 and 1998

22 CONCLUSION 100 m vertical difference corresponds to 35 mm increase of annual precipitation, with a decrease from the oceans towards the inner continent. The annual mean and the first three Fourier components explain 94 % of variance of the 25 years mean; but 66% of year-by-year anomalies. Response of precipitation is not unequivocal along the upper-Danube region. Hungary is characterised by slightly negative changes in both half-years. To the East in the Carpathians, the signs are un- equivocally negative, to the west in the Alpines they are positive. For an expected 0.5 K global warming the order of local changes is a few tens of percents of the total amount, in either direction. Precipitation of the independent 1999-2003 period supports the extrapolation of the observed negative trends in the 1974-1998 warming period for the Transsylvanean basin.

23 CONCLUSION (contd.) These changes are qualitatively supported by independent empirical and GCM-outputs, including precipitation and also cloudiness, however, the latter changes are rather diverse among the models, and somewhat smaller. Changes of cloudiness and OLR qualitatively remind the decreasing precipitation in the observed (by now) Eastern European region. Sea level pressure fields exhibit anticyclone–like relative modification in the central and eastern parts of the region, with precipitation decrease; whereas the observed precipitation increase in the Alpine region can not be directly related to the large scale sea-level pressure changes. Tendencies of runoff extremes may differ from those of preci- pitation, as illustrated on example of Transsylvanean basin.

24 (Instead of) DISCUSION - Change in thermal conditions! IPCC 2001 WG-I: Fig. 10.15: RegCM winter and spring warming increases with the altitude. A vertical snow- -albedo feedback? Alpine sub-region (Giorgi et al. 1997)

25 ADDITION TO A REMARK IN VITUKI (DECEMBER 2003) János MIKA How effective the diurnal circulation patterns are? How effective the diurnal circulation patterns (by Péczely for Hungary) are?

26 CIRCULATIONAL AND PHYSICAL FACTORS OF THE ANOMALIES Any local anomaly can be resolved (Mika, 1993) as: MM =  {q I }{  A I }+  {  A I }+ I=1 I=1 M +  {q I } +  I=1I=1 where the 1st term is zero.  = C+P+M C =  {  A I } part of due to anomalous frequency distribution of circulation types ( circulation term), P =  {q I } part of not directed by frequency of circ. types ( physical term), M =  part of due to mixed influence.

27 Results from the 13 Péczely macrosynoptic types (Molnar J. et al., 2001): Low portion of C term in * But fair correlation of case of macro-circulation. * C and full anomalies

28 Debrecen 1901-1996: Correlation Debrecen 1901-1996: Correlation of the clearly circulation term with the monthly anomaly, averaged. for 1, 4, 6, 8 years, and also for 12, 16 and 24 years.

29


Download ppt "PRECIPITATION TENDENCIES & DIAGNOSTIC PRECURSORS IN THE UPPER DANUBE CATCHMENT PARALLEL TO GLOBAL WARMING (1974-2003) János Mika 1, Vera Schlanger 1,"

Similar presentations


Ads by Google