THE CASE FOR MODIFIED GRAVITY James Binney Oxford University Outline MOND as a replacement for DM (Sanders & McGaugh 02) Absence of DM interior to the Sun (Bissantz et al 03, 04) TeVeS Lorentz-covariant MOND (Bekenstein 2004) NGC 3198 Begeman (1987) Modifying gravity Modify Newtonian theory at large distances? or at low accelerations? Adding a0 Bekenstein—Milgrom Eq. Tully-Fisher • Deep MOND regime – when µ(x)~x Sanders & Verheijen • At large r always enter deep MOND Fits to vc(r) for both LSB & HSB Galaxies (Sanders & McGaugh 02) a0=1.2 10-8 cm s-2 a0~H0c/2π; Λ~3(a0/c)2 U Maj Sanders & Verheijen Recover predicted M/L values Data: Sanders & Verheijen Models: Bell & de Jong 01 Giant E galaxies Data: Romanowsky et al 03 Models: Milgrom & Sanders 03 Solid: isotropic Clusters of Galaxies dSph galaxies η = Fi/Ft DM in the MW? Bissantz & Gerhard (02) Determine near-IR luminosity density from COBE K & L photometry Advances previous work by including spiral structure in disk Bissantz Englmaier & Gerhard (03) study gas flow in Φ obtained with spatially const M/L + quasi-isothermal DM halo Fit M/L, ωbar, ωspiral M/L for stars set by dynamics of nonaxisymmetric structure DM halo makes up balance for tangent-velocity curve Bissantz Englmaier & Gerhard CO observed simulated m=2 x 1.5 Bissantz Englmaier & Gerhard (03) Find ωbar in good agreement solar nhd kinematics With 4 arms get good pattern of ridge lines Vc near sun ~35 km/s below true value unless DM halo with a=10.7 kpc added With (x)=x/(1+x) KhN/KhM=0.95§0.15 (Famaey & B 04) Microlensing Optical depths Bissantz & Gerhard (02) Bissantz Debattista & Gerhard (04) Use novel N-body technique to find dynamical model that reproduces Bissantz & Gerhard photometry Adopt M/L, ω normalization from BEG No free parameters in Φ Reproduce proper motions of bulge stars in Baade’s window etc For plausible mass function of stars, reproduce MACHO microlensing event duration distribution (ML<,ML>)=(.04,10) or (.075,10) Conclusion: stars-only MW gives good fits to both optical depth & duration distribution Klypin et al (02) ΛCDM models of MW Adiabatic compression & optional L exchange No L exchange L exchange TeVeS Bekenstein (04) presents Lorentzcovariant theory (TeVeS) that reduces to MOND in appropriate limit Standard cosmologies Grav. Lensing as if DM present No superluminal modes TeVeS important development Link to effective field theory? Can now extend MOND to CMB and largescale structure If not worse than CDM in these fields, must be favoured theory Then question: significance of Uµ and Φ fields in TeVeS Conclusions MOND has amazing ability to model data taken after it was invented Excellent fits to galaxy rotation curves require M/L(colour) as from SS theory Compelling evidence that negligible DM interior to Sun Now limiting form of Lorentz covariant theory MOND really might be the next great step in physics