I think we don't need to set any source terms because they depend on the

gust derivatives, which are zero in all implemented cases. This piece of
code belongs to the split velocity method, which is currently not used
as far as I understand. However, when trying to remove it, I broke the
code elsewhere.

SU2_CFD/src/numerics/flow/flow_sources.cpp

@@ -784,8 +784,13 @@ CNumerics::ResidualType<> CSourceWindGust::ComputeResidual(const CConfig* config
     //residual[3] = smz*Volume;
     residual[3] = se*Volume;
   } else {
-    SU2_MPI::Error("You should only be in the gust source term in two dimensions", CURRENT_FUNCTION);
-  }
+	  //    SU2_MPI::Error("You should only be in the gust source term in two dimensions", CURRENT_FUNCTION);
+	  	residual[0] = 0.0;
+	  	residual[1] = 0.0;
+	  	residual[2] = 0.0;
+	  	residual[3] = 0.0;
+	  	residual[4] = 0.0;
+	    }
 
   /*--- For now the source term Jacobian is just set to zero ---*/
   //bool implicit = (config->GetKind_TimeIntScheme_Flow() == EULER_IMPLICIT);