Cloud–climate feedbacks are a major source of uncertainty in climate change projections. Satellites can monitor how clouds affect Earth's overall energy balance. But gaps in the data set owing to instrument failure substantially increase the amount of time needed to detect trends, finds a new study.
Norman Loeb of the NASA Langley Research Center in Virginia and colleagues simulated a 30-year record of cloud radiative effects — based on five years of satellite observations — and examined the impact of data gaps on the ability to detect long-term trends above natural variability. They show that the extent to which a data gap distorts trends depends on its positioning and length. A one-year gap in the middle of the record perturbs the trend more than a one-year gap at the beginning of the record, for example. Despite this variability, a gap of any length, anywhere in the record, prevents comparison of data collected before and after the gap because of potential calibration shifts, lengthening the time required to detect trends.
The researchers suggest that an overlap of at least six months between successive instruments is needed to avoid the adverse effects of gaps in the radiation record.