1. Branko Grisogono Geofizički odsjek, PMF Zagreb Hvala: Č. Branković & I. Güttler, DHMZ Glavne značajke klimatskih promjena i klimatskih scenarija

2. OUTLINE  Data: Global Past, Present → Future  Clearing out the Denial of Climate Change  Numerical Meteorological & Climate Models  Simulators  Climate vs. Weather:expect ≠ get  More of Current Results:Present → Future  Regional Climate Change  Tentative Conclusions → Discussion

3. OUTLINE  Data: Global Past, Present → Future  Clearing out the Denial of Climate Change  Numerical Meteorological & Climate Models  Simulators  Climate vs. Weather:expect ≠ get  More of Current Results:Present → Future  Regional Climate Change  Tentative Conclusions → Discussion

4. 5 th IPCC, late 2013  observed data

5. 5 th IPCC, late 2013  observed data, cont’d

6. Climate last ~ 140 yr: economy ↔ 2 x [economy] ↔ 2 x [CO2, CH4, aerosol, land-use,…] → 2 x [warming,…] ≈> Global CO2 as a linear surrogate for all the anthropogenic forcings b) Residues ≈ Actual T - Anthropogenic T Mean Residues ≈ ±0.11 o C  errors in 1-yr GCM hindcasts  Probability that it’s simply a giant natural fluctuation –how much rare?  Estimate the likelihood of a given amount of natural temp. change → bell curve  chance of a 1 o C fluct. over 125 yr. as natural is ≈ 1:10 5 or 1:3∙10 6  1 o C fluct. ↔ 5 σ  Nonlin. Geophys.→ extremes far stronger than from bell curve; maybe 100 x more likely → 1:1000; yet small enough to reject the possibility ≈2.33 o C /[2 x CO2] ↓

7. BASIC ISSUES: - MEASUREMENTS, WITH THEIR ERRORS, ARE SPARSE IN SPACE & TIME. THUS, INITIAL STATE OF THE ATMOSPHERE & OCEAN IS ONLY KNOWN PARTLY - NUMERICAL WEATHER PREDICTION (NWP) & CLIMATE MODELS CAN'T FULLY RESOLVE IN SPACE & TIME ALL THE RELEVANT PHYSICAL PROCESSES (ASIDE APPROXIMATIONS) - ATMOSPHERES & OCEANS ARE INHERENTLY CHAOTIC BESIDES STRONG DETERMINISTIC COMPONENTS

8. - The atmosphere’s behavior is governed by a set of physical conservation laws telling how the air moves (Newton’s laws of motion), about heating-cooling (laws of thermodynamics), roles of moisture… - The governing Partial Differential Equations (PDE’s) can’t be solved analytically – only numerical (i.e., approximate) solutions are possible * Given the current state, I.C. & B.C., the equations may be used to pass the info forward in time  forecast * The observations give the sparse current state – an incomplete picture, the weather & climate models may process the obs. → a fuller picture of reality

9. Unresolved Processes  Parameterized Parameterized: Radiation, Moist Processes, Turbulence,…

10. Predictability, like Turbulence, is Flow Dependent

11. Weather vs. Climate Models Resolution & integration length of the governing PDE’s (motion, mass, thermodynamics, spicies) Parameterizations (different space - & time-scales) e.g., weather models might have slight drift (may avoid some feedbacks, etc.) …climate modes not! Sometimes numerics, due to numerous couplings (feedbacks) being modeled differently, etc. Oceans, soil, biosphere, ice,… treated differently

12. 5 th IPCC, late 2013 - numerical simulations

13. Is the extreme weather we see today really caused by global warming? - Claims are made that push beyond what science can tell us. Attributing cause ↔ effect to individual weather events is fiendishly difficult. Climate is about patterns, statistical behavior… Extreme weather eventNo global warming linkMay change with global warming but amount not established Evidence of global warming link HurricanesX TornadoesX DroughtsX Forest firesX Heat wavesX Coastal floodsX EarthquakesX FloodsX

14. srednjak ansambla HadGEM1 GFDLCM21 EH5OM RegCM Temp. at 2m, Summer (2011-2040) - (1961-90) reanalysis obs.

15. Precipitation, winter (2011-2040) - (1961-90) reanalysis obs. HadGEM1 GFDLCM21 EH5OM Ensemble mean RegCM

16. Extreme Events No. Summertime Warm Days No. Wintertime Cold Days

17. Courtesy of Damir Počakal, DHMZ -There is a +trend in the mean duration of hailstone events in cont. Croatia, Počakal (2012) -But typical, standard climate models don't have those variables included yet - should be there! Average duration of hail episodes: continental Croatia: 4.3 min in 1981-2015, red & 6.3 min over the Polygon (NW Croatia), 2002-2015, blue. HAIL EPISODES IN CROATIA

18.  Current surface temp. upward trends & future predictions are unusual, unexplained by natural internal climate variability at ≈ 99 % confidence  “Predictions” based on greenhouse gas emission scenarios & links the gases concentrations  temp. via modeling & simulations  Included: population growth, economic development, technol. change, social interactions  Besides large spatio-temporal variations, global warming is real, it exceeds in magnitude & pace natural changes over more than the last 10 3 yr.  No detailed clim. projections for hail-storms and wind fields in/around Croatia TENTATIVE CONCLUSIONS

19.  Immediate changes seem needed: lowering emissions of green- house gasses to meet the scenario with global ~ +2 o C in ≈ yr. 2100  Targeting and promoting more human technologies, renewable energies, healthier food & water  Humanistic approach needed!  It is wrong to deploy “instrumental rationalism” (max. efficiency only) & blind pragmatism based on e.g., large resources & markets  Make reliable regional climate – economy projections & space- time variability for next few decades (agronomy, energy, tourism, education, traffic, etc.) TENTATIVE CONCLUSIONS cont’d

20. The End

21. IPCC 2007 Model Simulations