Elsevier

Journal of Hydrology

Volume 513, 26 May 2014, Pages 467-474
Journal of Hydrology

Nowcasting of rain events using multi-frequency radiometric observations

https://doi.org/10.1016/j.jhydrol.2014.03.066Get rights and content

Highlights

  • Microwave radiometric observations are used to nowcast heavy rain events.

  • Brightness temperatures of 22.24 GHz and 58 GHz show concurrent change before rain.

  • Brightness temperature derivatives indicate favorable conditions for imminent rain.

  • The prediction technique generates alarm 25 min before rain for 90% of cases.

Summary

Nowcasting of heavy rain events using microwave radiometer has been carried out at Kolkata (22.65°N, 88.45°E), a tropical location. Microwave radiometer can produce the temperature and humidity profiles of the atmosphere with fairly good accuracy. Definite changes are observed in temperature and humidity profiles before and at the onset of heavy rain events. Concurrent changes in the brightness temperatures (BT) at 22 GHz and 58 GHz are found to be suitable to nowcast rain. The time derivatives of brightness temperatures at 22 GHz and 58 GHz are used as inputs to the proposed nowcasting model. In addition, the standard deviation of the product of these time derivatives is also considered. The model has been developed using the data of 2011 and validated for rain events of 2012–2013 showing a prediction efficiency of about 90% with alarm generated about 25 min in advance.

Introduction

In the Indian subcontinent, sudden and heavy precipitations occur during pre monsoon period of March–May and monsoon period of June–September. These forms of precipitations are found to affect agriculture, aviation and in severe cases can cause loss of life and property. Nowcasting of heavy precipitations has many applications in mitigating some of the adverse situations resulting from such events.

Conventionally, satellite and radar data are used to nowcast thunderstorms (Browning, 1982, Cluckie and Collier, 1991, Dutta et al., 2010, Mecklenburg et al., 2000, Sokol, 2006, Wang et al., 2009, Wilson et al., 1998, Zahraei et al., 2013). The reported prediction efficiencies are usually around 80% with false alarm rates above 25% for both radar and numerical weather prediction (NWP) based nowcasting (Johnson and Olsen, 1998, Wilson et al., 1998, Lin et al., 2005). The temperature and humidity profiles of the atmosphere along with the instability indices obtained from atmospheric sounding measurements such as radiosondes can also be a useful tool for rain and thunderstorm predictions (Geerts, 2001, Manzato, 2003, McCann, 1994). Except radar, these instruments suffer from poor temporal resolution. However radars are costly and need much involved maintenance for continuous operation.

Heavy rain events also create attenuation of microwave signals. So microwave propagation through the atmosphere can provide a useful signature of heavy rain events. In this connection atmospheric brightness temperatures measured by a microwave radiometer can be useful for nowcasting of rain (Koffi et al., 2007, Won et al., 2009). Güldner and Spänkuch (1999) showed that the liquid water content (LWC) and perceptible water vapor (PWV) increases before rain. So an increase of brightness temperature (BT) in water vapor channels (22–31 GHz) can be observed about 2 h before rain (Won et al., 2009). Many convective indices from radiometer have been utilized for nowcasting heavy precipitation events (Darkow, 1968, Faubush et al., 1951, Galway, 1956, Madhulatha et al., 2013, Showalter, 1953). There are some efforts to predict rain using a microwave radiometer. Dvorak et al. (2012) used BT at 10 GHz to predict rain with a hit ratio of 74% and a false alarm rate of 7%. Won et al. (2009) also used BT at 22, 30 and 51 GHz for rain prediction with a probability of detection 0.9 for rain accumulated below 20 mm.

However, models using radiometric observations are not highly accurate for nowcasting rain (Chan and Lee, 2011, Wilson et al., 1998, Won et al., 2009). The primary reason could be that an increase of atmospheric water vapor has been taken to be the only precursor of intense convective activities. However, water vapor can also increase significantly in the absence of rain (Sherwood et al., 2010, Won et al., 2009). So monitoring of other parameters of the atmosphere is also needed.

The radio environment over Kolkata (22.65°N, 88.45°E) has been studied using a multi-frequency profiler radiometer. Since convective rain is usually associated with drastic change in water vapor as well as temperature, brightness temperatures of water vapor and oxygen absorption bands can be considered to be indicators of the impending rain events.

Section snippets

Experimental setup and data

A Dicke radiometer system (RPG-HATRO) is used for the present study. It consists of two receiving sections along with a noise diode, a data acquisition system, rain sensor, GPS clock, and a ground pressure and temperature sensor (Rose and Czekala, 2009). It measures the brightness temperatures in the range of 0–800 K with an accuracy of 0.5 K at 14 frequencies in two frequency bands (7 frequencies in each band). As the first frequency band (22–31.4 GHz) is sensitive to water vapor absorption, it

Methodology

During the developing stage of convection, the warm air rises up causing an updraft. The entrainment of warm air causes cold air from the surroundings to fill the region, thereby producing a cooling effect. This stage occurs before rain occurrence and it usually lasts for about 10–15 min (Byers and Braham, 1948, Rogers and Yau, 1982). Also, during the mature stage, a liquid mass, too heavy to be sustained as cloud, comes down in the form of rain generating precipitation drag and evaporative

Validation

To validate the efficacy of the model, it is tested on a new dataset comprising 130 rainy days of premonsoon and monsoon period of 2012 (March 2012–September 2012) and premonsoon season of 2013 (March 2013–June 2013). The actual prediction efficiency obtained is about 90% while the over prediction is 11%.

The time difference between event occurrences and alarm generation is recorded for all events. When the frequency distribution of these time gaps is plotted, it is seen that the alarm is mostly

Conclusions

In this paper it has been observed that the brightness temperature at 22 and 58 GHz, which are sensitive to humidity and temperature profiles respectively, have shown sharp changes before rain. So, the simultaneous variations of brightness temperatures at 22 and 58 GHz have been utilized to nowcast heavy precipitations. The present technique involves radiometric observations in the water vapor as well as the oxygen absorption band which has not yet been reported in open literature. A model has

Acknowledgement

The financial support provided by Indian Space Research Organisation (ISRO), India under the projects (1) “Integrated studies on water vapor Liquid Water content and Rain of Tropospheric atmosphere and effects in Radio Environment” and (2) “Space Science Promotion Scheme” are thankfully acknowledged.

References (32)

  • G.L. Darkow

    The total energy environment of severe storms

    J. Appl. Meteorol.

    (1968)
  • Das, S., Chakraborty, R., Talukdar, S., Maitra, A., 2012. Nowcasting of tropical rain using dual frequency atmospheric...
  • D. Dutta et al.

    Nowcasting of yes/no situation at a station using soft computing technique to the radar imagery

    IJRSP

    (2010)
  • P. Dvorak et al.

    Short-term prediction and detection of dynamic atmospheric phenomena by microwave radiometer

    Radioengineering

    (2012)
  • E.J. Faubush et al.

    An empirical method for forecasting tornado development

    Bull. Am. Meteorol. Soc.

    (1951)
  • J.G. Galway

    The lifted index as a predictor of latent instability

    Bull. Am. Meteorol. Soc.

    (1956)
  • Cited by (40)

    • Aliasing effect due to convective rain in Doppler spectrum observed by micro rain radar at a tropical location

      2018, Advances in Space Research
      Citation Excerpt :

      In stratiform rain the wind drag effect is not completely absent, but, according to Tridon et al. (2011), the wind draft impact is significant at heights greater than 300 m above ground level. The present study location Kolkata experiences frequent convective rain and thunderstorm activities accompanied by strong vertical wind in the pre-monsoon and monsoon period (Chakraborty et al., 2014; Rao et al., 2001; Thurai et al., 2003; Rakshit et al., 2016; Market et al., 2017; Rakshit et al., 2017). This causes Doppler shift of the radar signal beyond the unambiguous measurement range and consequently an aliasing error creeps in the Doppler spectrum (Tridon et al., 2011).

    • Can Nor'wester events initiate stratospheric moistening?

      2017, Journal of Atmospheric and Solar-Terrestrial Physics
      Citation Excerpt :

      A decreasing OLR from MERRA-2 dataset is seen during Nor'wester event on 8 May 2014 when compared to other 10 cal m days of the same month (Fig. 2). The strength of a typical convective event can be characterized by its Convective Available Potential Energy (CAPE), representing the amount of buoyant energy available to speed up a parcel vertically from the Lifted Condensation Level (LCL) to its equilibrium level (EL) (Chakraborty et al., 2014, 2015, 2016, 2017b; Maitra et al. 2014; Saha et al., 2014). A high CAPE value can thus be held responsible for lifting of water vapour to the UTLS region resulting in stratospheric moistening.

    • Association of atmospheric pollution and instability indices: A detailed investigation over an Indian urban metropolis

      2017, Atmospheric Research
      Citation Excerpt :

      Hence, studies related to atmospheric convective processes have earned a great importance in the past few decades over the globe (Sun and Wang, 2008; Horváth et al., 2009; Midya et al., 2011; Saha and Maitra, 2014; Saha et al., 2012, 2014, 2017b; Melani et al., 2013; Ullah and Shouting, 2013; Guo et al., 2014; Chakraborty et al., 2015; Chakraborty et al., 2016). Both long and short-term variations in atmospheric processes before and during such convective events have also been investigated over different parts of the world (Kašpar and Müller, 2007; Sokol and Pešice, 2009; Midya and Saha, 2011a; Gatzen, 2013; Chakraborty et al., 2014, 2016; Saha et al., 2014; Chen et al., 2016; Bližňák et al., 2017). Atmospheric instability plays the most significant role in the transport and dispersion of air pollutants.

    View all citing articles on Scopus
    View full text