not as straightforward as it seems at first glance 
Lets consider the case of one turning point only (all other cases can be shown equivalent to it
)->
In this also , lets make 2 cases
Case 1 ->
In this case, lets consider the most symmetric case , ie when it turns at t=1/2 sec and it has same acceleration throughout
At t=1/2
s1=at^2 /2 =a/8
v1=at=a/2
t=1/2 to t=1
v2= 0 = u -At =a/2 - At =a/2- A/2
implies a = A
s2=ut - at^2/2 =a/4 - a/8 =a/8
but s1 + s2 = 1 = a/4
implies a = 4 m/s^2
the velocity time graph in this case is as shown in graph 1
------------------------------------------------------------------------------------------------------------------------------------------------------
Case 2- >
Now , suppose instead of t=1/2 , the turning point comes at some other time , then it is obvious that if acceleration is less than 4 for interval before turning point , acceleration would be greater than 4 for for interval after turning point and vice versa
------------------------------------------------------------------------------------------------------------------------------------------------------
Now , we have to proove that instead of 1 turning point suppose there are infinitely many turning points yet the above condition holds true .
For it , consider the graph 2 in which there are many turns .
Now we can consider any point P , and then we can see the the part before P will correspond to a certain average acceleration and part after P will correspond to a certain average acceleration , and hence the corresponding graph would come to be of either case 1 or case 2.
Now suppose , before P ,the average acceleration is greater than 4 , then since there are many turning points its obvious that to maintain average acceleration of more than 4 , at some points instantaneous accelerations would have to be greater than 4 .Same would be true for interval after P , if it has average acceleration greater than 4.
Hence , option a and c are correct