When working with IO (Input-output) model for sectoral CO2 calculation sometimes you have negative 'changes in inventory' sometimes big enough to make total final demand for that sector negative. which will obviously undermine the actual CO2 emissions for that sector. if due to this negative changes in inventory we calculate a sectors total emissions to be negative for a particular year what would be the best interpretation for such a negative result?

I mean how can I present this kind of result?

or we can ignore changes in inventory when working with IO tables?