Unexpected result 'UNSAT' for sat-formula

[kensingRichardt]

Hello again,

I found a case where eldarica returns unexpectedly UNSAT:

(declare-datatypes ((k 0)) (((k (y Int)))))
(declare-datatypes ((e 0)) (((e (f (Array Int k))))))
(declare-datatypes ((g 0)) (((g (h (Array Int e))))))
(declare-fun j (g) Bool)

(assert (forall ((r g))
    (=>  (> (y (select (f (select (h r) 0)) 0)) 0) (j r))))
   

(assert (forall ((r g))
    (=  (> (y (select (f (select (h r) 0)) 0)) 0) (j r))))
    
(check-sat)

Here is how I invoked eldarica:

$ ~/eldarica/eld -cloneArrays sat.smt2 
unsat

When omitting option -cloneArrays we get sat as expected.
I did not expect that this option changes the sat-result or is this intended behavior?