The Principle of Satellite Remote Sensing Precipitation

introduction

Precipitation is all water, solid and liquid, that falls from clouds to the Earth's surface, mainly in the form of rain and snow.

In April 1960, since the launch of the first meteorological satellite—Television and Infrared Radiation Observation Satellite 1 (TIROS-1)—people have obtained a large amount of space remote sensing information, which provides the possibility to retrieve precipitation. Early remote sensing precipitation retrieval mainly relied on passive remote sensing, including visible light (visible, VIS), infrared (infrared, IR) and active/passive microwave (active/passive microwave, AMW/PMW).

Visible light and infrared sensors can be carried on satellites in geosynchronous orbit (GEO) and low earth orbit (LEO), while microwave sensors can only be carried on satellites in low earth orbit.

Geostationary and low-Earth orbit satellites

1. Visible light and infrared sensors of geostationary satellites usually observe the target area every tens of minutes, with high time resolution, which can provide satellite cloud images and capture some precipitation cloud systems with a short life history.
2. The sensors carried by low-Earth orbit satellites will have blind spots when scanning, but the satellite cloud images provided by the microwave channel can effectively reduce the impact of cirrus clouds on the accuracy of precipitation retrieval.

Visible light, infrared and active and passive microwave

1. Visible light and infrared: higher spatial and temporal resolution, which are indirect estimates.
2. Passive microwave: strong physical significance, more accurate calibration, quite flexible frequency band selection, easy to operate, brightness temperature is sensitive to hydrometeors.
3. Active microwave (spaceborne radar): the calibration is quite accurate and the reflectivity is measured; the scan width is narrow and the test is strong.

different algorithms

Visible light and infrared precipitation retrieval algorithm

Visible light and infrared precipitation inversion algorithm is the earliest and the simplest method, which takes advantage of the physical properties of cold and warm clouds. The presence of cold and warm clouds is associated with convection, which produces precipitation. Cloud top information measured in the visible and infrared bands can be used to indirectly estimate surface precipitation. Specifically, the relationship between cloud top infrared temperature and precipitation probability and intensity is established.

The geostationary satellite GEO provides long-term, relatively continuous monitoring data in the visible and infrared bands, which can provide very fine information on changes in precipitation intensity.

microwave (passive)

The microwave precipitation retrieval algorithm has three main advantages. First, when precipitation is observed by passive microwave, raindrops will strongly affect the microwave radiation transmission process, so the spaceborne microwave radiometer can easily detect precipitation information. Second, microwaves are highly penetrable in cloudy and rainy atmospheres, and can work around the clock under severe weather conditions. Third, the radiation information generated inside the precipitation cloud can reach the spaceborne microwave radiometer, so it already contains the information of the precipitation structure, so the microwave precipitation inversion is a more direct inversion algorithm with a more solid physical foundation .

Land and ocean have their own different microwave radiation characteristics. For the ocean, the microwave specific emissivity of the sea surface is low, so the background radiation information of passive remote sensing is small and close to a constant. In this context, the emitted radiation signal of precipitation is strong, coupled with the difference between the low-polarization characteristics of precipitation and the high-planarity characteristics of the sea surface, so the sea surface precipitation can be identified and quantitatively retrieved in the low frequency band. For land, the microwave specific emissivity at the ground is usually high and varies widely, so the emitted radiation from hydrometeors cannot be well identified and quantified. At the same time, the polarization characteristics of the land surface are not obvious, which increases the difficulty of inversion of land precipitation. The scattering effect of ice particles in the high-frequency band will weaken the upward radiation intensity on the surface, which can be used to retrieve land precipitation.

Microwave sensors are only carried on low-earth orbit satellites , so passive microwave algorithms are only applicable to this type of spaceborne sensors. The ocean space resolution is about 50km, and the land resolution is about 10km.

Precipitation Radar Retrieval Algorithm

The TRMM is equipped with the first active microwave sensor (precipitation radar) dedicated to the detection of precipitation. It is a phased array weather radar that mainly uses the 13.8GHz frequency band to observe precipitation particles and reflected energy on the earth's surface, and can obtain three-dimensional spatial structure information of ocean and land precipitation.

However, this approach has some disadvantages. First of all, its observation range is limited; at the same time, it has the weakness of ground-based radar; attenuation correction and precipitation estimation methods of radar observation data are also affected by parameter uncertainties. Contributing factors include: raindrop spectrum, precipitation particle phase, density and shape, inhomogeneous distribution of precipitation within imaging radar pixels, attenuation of radiation intensity due to cloud liquid water and water vapor, freezing height, scattering cross-section uncertainty, and radar Echo signal changes, etc.

joint algorithm

The above three methods have their own advantages and disadvantages, so the current study considers joint use to maximize the quality of precipitation observations. Next, some representative methods are introduced:

CMORPH deformation algorithm

The algorithm uses GEO satellite infrared data every half hour to interpolate the PMW retrieval data, so as to obtain relatively fine precipitation intensity in space and time. The spatial and temporal distribution of precipitation depends on PMW, not on the numerical value of infrared data.

TPMA

TPMA has the following four steps:

1. Calibration and fusion of microwave precipitation estimates;
2. Generate visible/IR precipitation estimates using calibrated microwave precipitation estimates; (interpolation?)
3. Fusion of microwave precipitation estimation and infrared precipitation estimation;
4. Fusion of surface rain gauge observation precipitation data.

GSMaP

The algorithm uses various attributes of TRMM data, and uses precipitation radar inversion algorithm, rain/no-rain classification method and scattering algorithm to estimate the hydrometeor profile.

summary

Visible light and infrared algorithms are an indirect method with limited accuracy, but infrared data has a high time sampling frequency and a large coverage area; passive microwave (mounted on near-Earth satellites) is a direct method, and its algorithm accuracy is higher than that of visible light and infrared Algorithm, but the algorithm is complex and the data temporal and spatial resolution is low; the precipitation radar inversion algorithm has the highest accuracy, and can provide three-dimensional spatial precipitation structure, but its sampling frequency and coverage are small. Therefore, it is a research frontier to monitor and retrieve precipitation by means of multi-satellite, multi-channel, and multi-model joint means.

appendix

electromagnetic spectrum

 Atmospheric window : When sunlight passes through the atmosphere, it will be affected by the absorption and scattering of sunlight by the atmosphere , thus attenuating the energy of sunlight passing through the atmosphere. But the absorption and scattering of sunlight by the atmosphere varies with the wavelength of sunlight. Atmospheric windows are some wave bands that can penetrate the atmosphere in celestial radiation, mainly including: microwave band, thermal infrared band, mid-infrared band, near ultraviolet, visible light and near infrared band.

​

Electromagnetic spectrum diagram. The picture comes from Weibo

​

Classification of infrared light. The picture comes from the blog garden

references

[1] Liu Yuanbo, Fu Qiaoni, Song Ping, Zhao Xiaosong, Dou Cuicui. A review of satellite remote sensing retrieval of precipitation [J]. Advances in Earth Sciences, 2011, 26(11): 1162-1172.

[2] Tang Guoqiang, Wan Wei, Zeng Ziyue, Guo Xiaolin, Li Na, Long Di, Hong Yang. Overview of Global Precipitation Measurement (GPM) Program and Its Latest Progress [J]. Remote Sensing Technology and Application, 2015,30(04):607 -615.

[3] Wang Cunguang, Hong Yang. Summary of Inversion, Verification and Application of Satellite Remote Sensing Precipitation [J]. Water Conservancy and Hydropower Technology, 2018, 49(08): 1-9. DOI: 10.13928/j.cnki.wrahe.2018.08. 001.