The increasing of the accuracy of the measurement systems of satellites for optic remote sensing in recent years allows solving fundamentally new problems connected with determining of the global distribution of the small gas components in the atmosphere. All of such measurements are indirect. Its leads to stiffening of the requirements to the accuracy of solving radiative transfer equation (RTE). At first it is necessary to include in the model all the factors which affect the signal significantly. Every moment at least half of the planet is covered by clouds. Due to this factor measurements almost always are making in breaks in cloudiness. Clouds affect the signal significantly so one should consider it in modeling. From the point of view of inverse problems high speed of the algorithm is very important too. There are no algorithms which satisfy fully the requirements of the inverse problems for now. In this paper the new method of RTE solving based on quasi-twostream approximation was proposed. The method is a combination of the approximate method with one iteration to adjust radiance angle distribution (synthetic iterations). As an approximate method we use twostream approximation – the simplest and the fastest method of solving RTE. We consider the cylindrical hole in a plane parallel cloud as a case of broken cloudiness.