Ranked set sampling(RSS) designs are used when the units can be easily ranked with respect to the variable of interest or the correlated auxiliary variable. In this study we have take in to account RSS and median ranked set sampling (MRSS) designs with stratification for population mean. We have examined estimators both simple random sampling and stratified random sampling and extended these estimators for stratified RSS and stratified MRSS. We have conducted a simulation study using a real data set. With this data set we have tried to estimate air-quality in Turkey using different sampling designs and estimators. The simulation study showed that we obtained most efficient results under stratified median ranked set sampling design.