Underestimation Of Rainfall Research Articles

Evaluation of High-Resolution Satellite Precipitation Products over Very Complex Terrain in Ethiopia

Abstract This study focuses on the evaluation of 3-hourly, 0.25° × 0.25°, satellite-based precipitation products: the Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA) 3B42RT, the NOAA/Climate Prediction Center morphing technique (CMORPH), and Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN). CMORPH is primarily microwave based, 3B42RT is primarily microwave based when microwave data are available and infrared based when microwave data are not available, and PERSIANN is primarily infrared based. The results show that 1) 3B42RT and CMORPH give similar rainfall fields (in terms of bias, spatial structure, elevation-dependent trend, and distribution function), which are different from PERSIANN rainfall fields; 2) PERSIANN does not show the elevation-dependent trend observed in rain gauge values, 3B42RT, and CMORPH; and 3) PERSIANN considerably underestimates rainfall in high-elevation areas.

Calibration of TRMM Derived Rainfall Over Nepal During 1998-2007

In this study rainfall is calculated from Tropical Rainfall Measuring Mission (TRMM) Version 6 (V6) 3B42 datasets and calibrated with reference to the observed daily rainfall by rain-gauge collected at 15 locations over Nepal during 1998-2007. In monthly, seasonal and annual scales TRMM estimated rainfalls follow the similar distribution of historical patterns obtained from the rain-gauge data. Rainfall is large in the Southern parts of the country, especially in the Central Nepal. Day-to-day rainfall comparison shows that TRMM derived trend is very similar to the observed data but TRMM usually underestimates rainfall on many days with some exceptions of overestimation on some days. The correlation coefficient of rainfalls between TRMM and rain-gauge data is obtained about 0.71. TRMM can measure about 65.39% of surface rainfall in Nepal. After using calibration factors obtained through regression expression the TRMM estimated rainfall over Nepal becomes about 99.91% of observed data. TRMM detection of rainy days is poor over Nepal; it can approximately detect, under-detect and over-detect by 19%, 72% and 9% of stations respectively. False alarm rate, probability of detection, threat score and skill score are calculated as 0.30, 0.68, 0.53 and 0.55 respectively. Finally, TRMM data can be utilized in measuring mountainous rainfall over Nepal but exact amount of rainfall has to be calculated with the help of adjustment factors obtained through calibration procedure. This preliminary work is the preparation of utilization of Global Precipitation Measurement (GPM) data to be commencing in 2013.

Application of multimodel ensemble techniques for real time district level rainfall forecasts in short range time scale over Indian region

During the summer monsoon (1 June to 30 September) 2007, real-time district level rainfall forecasts in short-range time scale were generated for Indian region applying multimodel ensemble technique. The pre-assigned grid point weights on the basis of correlation coefficients (CC) between the observed values and forecast values are determined for each constituent model at the resolution of 0.5° × 0.5° utilizing two seasons datasets (1 June to 30 September, 2005 and 2006), and the multimodel ensemble forecasts (day 1 and day 2 forecasts) are generated at the same resolution on a real-time basis. The ensemble forecast fields are then used to prepare forecasts for each district taking the average value of all grid points falling in a particular district. In this paper we examined the performance skill of the multimodel ensemble-based real-time district level short-range forecast of rainfall. It has clearly emerged from the results that the multimodel ensemble technique reported in this study is superior to each ensemble member. District wise performance of the ensemble rainfall forecast reveals that the technique, in general, is capable of providing reasonably good forecast skill over most districts of the country, particularly over the districts where the monsoon systems are dominant. Though the procedure shows appreciable skill to predict occurrence or non-occurrence of rainfall at the district level, it always underestimates rainfall amount, particularly in heavy rainfall events. Possible reasons of this failure may be due to model bias and poor data assimilation procedure.

Validation of western and eastern Pacific rainfall estimates from the TRMM PR using a radiative transfer model

Rainfall estimates from the Tropical Rainfall Measuring Mission (TRMM) Precipitation Radar (PR) facility algorithm over the eastern Pacific are lower than those from the TRMM Microwave Imager (TMI) facility algorithm and a PR‐consistent advanced microwave radiometer algorithm, while they are in good agreement over the western Pacific. We investigate the consistency between TMI‐observed brightness temperatures (TBs) for the 19 GHz channel and those simulated from the PR rainfall estimates using a radiative transfer model. Discrepancies between observed TBs and simulated TBs from the PR are larger for the eastern Pacific than for the western Pacific, indicating that PR underestimates rainfall over the eastern Pacific. It is hypothesized that drop size distributions (DSDs) in the eastern Pacific have stronger maritimity (i.e., more small to medium sized raindrops) than the initial DSD model of the PR algorithm, representative of the “data‐rich” western Pacific, leading to underestimation of rain by the PR. Differences in the adjustable parameter of the PR algorithm, implying changes in the DSD, and the vertical structure of PR‐observed reflectivity over the western and eastern Pacific support the hypothesis.

Resolving orographic rainfall on the Indian west coast

AbstractWe discuss a method to obtain the spatial estimates of rainfall on the Indian west coast for calculating the surface water budget. The region includes the basin of the Mandovi River, a typical west‐coast river whose catchment area includes the Sahyadris range. The rainfall maps obtained by a multivariate interpolation scheme, with elevation as the third variable, were able to capture the sharp increase in rainfall on the windward slopes and the exponential decay in rainfall on the leeward side. The rainfall maps were used to force a hydrological model and the simulated discharge was within 16% of the observed discharge in 16 of 18 years during 1981–1998; the average error over the 18 years was 8%, which is much less than the estimated error of ∼15–20% in the observed discharge. The Mandovi river discharge provided a reliable and independent validation of the accuracy of the rainfall estimate as the river is purely rain‐fed. The method provides a better estimate of rainfall in the Mandovi basin, and therefore for the Indian west coast, than existing rainfall datasets. The key result is that treating the windward and leeward sides separately by an a priori, subjective extraction of the ridge line can reduce the underestimation of rainfall that is common in mountainous terrain when the data coverage is sparse. Copyright © 2007 Royal Meteorological Society

Radar and Rain Gauge Analysis of the Extreme Rainfall during Hurricane Danny’s (1997) Landfall

Abstract As a minimal hurricane, Danny moved over Mobile Bay around 0900 UTC 19 July 1997 and became stationary by midmorning, while situated within a synoptic col. Danny then evolved into an asymmetric storm with an intensely convective rainband that produced torrential rainfall through 1200 UTC 20 July 1997. Danny’s center remained <100 km from the National Weather Service (NWS) Weather Surveillance Radar-1988 Doppler (WSR-88D) in Mobile, Alabama, for over 48 h, allowing long-term surveillance of the storm’s inner core. This event marked the first time the tropical Z–R relationship was employed on an operational WSR-88D system during tropical cyclone landfall. A radar-estimated maximum rainfall accumulation of 1097 mm (43.2 in.) was analyzed over southwestern Mobile Bay. A NWS cooperative rain gauge located on Dauphin Island, Alabama, measured 896 mm (35.28 in.). An adjacent standard rain gauge measured the highest rainfall amount of 932 mm (36.71 in.). This paper investigates the spatial and temporal distribution and potential magnitude of Danny’s torrential rainfall episode over coastal Alabama. It is shown that both gauges and radar seriously underestimated event rainfall. An estimate is given for what could have been the true event rainfall amount. In the case of the radar, the WSR-88D Algorithm Testing and Display System is used to obtain a better estimate of rainfall using higher dBZ caps than the operational 50 dBZ. In the case of the tipping-bucket rain gauge, wind and mechanical error estimates were applied in order to quantify rainfall underestimation.

Long‐term flow forecasts based on climate and hydrologic modeling: Uruguay River basin

This paper describes a procedure for predicting seasonal flow in the Rio Uruguay drainage basin (area 75,000 km2, lying in Brazilian territory), using sequences of future daily rainfall given by the global climate model (GCM) of the Brazilian agency for climate prediction (Centro de Previsão de Tempo e Clima, or CPTEC). Sequences of future daily rainfall given by this model were used as input to a rainfall‐runoff model appropriate for large drainage basins. Forecasts of flow in the Rio Uruguay were made for the period 1995–2001 of the full record, which began in 1940. Analysis showed that GCM forecasts underestimated rainfall over almost all the basin, particularly in winter, although interannual variability in regional rainfall was reproduced relatively well. A statistical procedure was used to correct for the underestimation of rainfall. When the corrected rainfall sequences were transformed to flow by the hydrologic model, forecasts of flow in the Rio Uruguay basin were better than forecasts based on historic mean or median flows by 37% for monthly flows and by 54% for 3‐monthly flows.

Rainfall Measurement Error by WSR-88D Radars due to Variations inZ–RLaw Parameters and the Radar Constant

Abstract An investigation is made of the extent to which variations in Z–R (reflectivity factor–rainfall rate) relations can explain the systematically large offsets of radar-measured rainfall from rain gauge measurements as observed with some National Weather Service (NWS) WSR-88D radars. It is shown that theory predicts that the use by the NWS of the current default Z–R relation (Z = 300R1.4) should underestimate rainfall by about 25% in stratiform rain and overestimate it by about 33% in thunderstorm rain. Yet it is commonly observed that some WSR-88D radars systematically underestimate rainfall by a factor of 2 or more in stratiform rain and produce estimates of rainfall with acceptable accuracy for isolated thunderstorms. It is shown that variations in Z–R law parameters alone cannot explain these discrepancies, and it is proposed that their origin lies in a radar calibration offset. A comparison is made of reflectivity factors measured by the KGSP WSR-88D at Greer, South Carolina, with those determi...

Mesoscale Precipitation Fields. Part I: Statistical Analysis and Hydrologic Response

Abstract A statistical objective analysis (SOA) scheme was developed to adjust estimates of rainfall accumulation from the WSR-88D in central Oklahoma using rain gauge measurements from the Oklahoma Mesonetwork. Statistical parameters of these rainfall fields were obtained from a time series of 2-km Constant Altitude Plan Position Indicators at 2 km × 2 km resolution for accumulation time intervals of 15, 30, 60, and 120 min. Results document that the Twin Lakes WSR-88D underestimates rainfall rates by 28% on average. In turn, these errors generate large errors in the streamflow simulations performed for the Dry Creek watershed in north-central Oklahoma. Mean normalized expected error variances of the accumulated rainfall over 1–2-h duration were reduced up to 25% by the SOA scheme. These results are discussed in the context of a hydrometeorological forecast system that uses the analyzed rainfall field to adjust rainfall rates for nowcasting (0–3 h), to improve rainfall forecasts (0–12 h) via its assimila...

An Intercomparison Study of NEXRAD Precipitation Estimates

Systematic biases in WSR‐88D (Weather Surveillance Radar–1988 Doppler) hourly precipitation accumulation estimates are characterized from analyses of more than 1 year of WSR‐88D data and rain gage data from the southern plains. Biases are examined in three contexts: (1) biases that arise from the range‐dependent sampling of the WSR‐88D, (2) systematic differences in radar rainfall estimates from two radars observing the same area, and (3) systematic differences between radar and rain gage estimates of rainfall. Range‐dependent biases affect hourly rainfall accumulations products over much of the area covered by the WSR‐88D. Significant underestimation of rainfall occurs within 40 km range of the radar due to bias in reflectivity observations at the higher elevation angles used for rainfall estimation close to the radar. Bright band and anomalous propagation (AP) lead to systematic overestimation of rainfall at intermediate range. Beyond 150 km in spring‐summer and beyond 100 km in winter‐fall, underestimation of precipitation is pronounced due to incomplete beam filling and overshooting of precipitation. Radar‐radar intercomparison studies suggest that radar calibration is a significant problem at some sites. Anomalous propagation during clear‐air conditions, a major problem with previous National Weather Service network radars, has been largely eliminated by the WSR‐88D processing. AP remains a problem for cases in which AP returns are embedded in rain. Radar–rain gage intercomparison analyses indicate systematic underestimation by the WSR‐88D relative to rain gages for paired gage‐radar rainfall estimates. Analyses of spatial coverage of heavy rainfall, however, illustrate fundamental advantages of radar over rain gage networks for rainfall estimation.

The Application of Radar–Gauge Comparisons to Operational Precipitation Profile Corrections

Abstract The analyses of data recorded during the past eight years with two Swiss radars, a network of rain gauges, and river flow measurements have helped to quantify the vertical profile of reflectivity and the influences of topography, meteorology, and radar parameters on the precision of radar precipitation estimation. The influence of the topography around the radar, the width of the radar beam, and the vertical echo structure produces a complex error distribution in space and time, with errors dependent upon storm type, distance from the radar, and the radar horizon. In spite of excellent agreement between amounts estimated by the 5-cm radar at close ranges and gauges located below the radar volume, underestimation of rainfall increases with range from the radar. The authors' experience dramatically shows how significantly errors are reduced when precipitation can be estimated close to the ground, a task made easier by choosing a radar site with a good view and by rigorously eliminating echoes conta...

The underwater sound generated by heavy rainfall

The underwater acoustic signature of heavy rainfall is very different from that of light rainfall. During heavy rainfall sound levels are observed to rise with increasing rainfall rate at all frequencies monitored (4–21 kHz) and the 15-kHz spectral peak observed during light rainfall is absent. The sound levels are most highly correlated (r≊0.8) with heavy rainfall rate for frequencies less than 10 kHz. Lower correlations between sound levels and heavy rainfall rate were observed for frequencies above 10 kHz under several different conditions. When wind speed exceeds 10 m/s, wave breaking mixes bubbles downward and creates a layer of bubbles. This bubble layer attenuates subsequent surface-generated sound (from the raindrop splashes) for frequencies above 10 kHz. Extremely heavy rainfall (total rainfall above 150 mm/h) also generates a subsurface bubble layer. This rainfall-generated bubble layer is evidence of rainfall-induced turbulent mixing of the ocean surface layer and has implications for air/sea exchange processes (momentum, heat, and gas exchange). Finally, previous studies have shown that light rain generates acoustic energy above 10 kHz and that this sound is poorly correlated with total rainfall rate. A simple empirical acoustic rainfall rate algorithm for heavy rain is offered. This algorithm may be site specific. Furthermore, it overestimates rainfall rate early in the rain events examined here (convective rain events), and then underestimates rainfall rate later in the same events. This observation is shown to be consistent with likely changes in the drop size distribution during the lifetime of the rain event. The sound produced by large drops within heavy rain (drop diameter greater than 2.2 mm) is shown to dominate the underwater sound field. The empirical acoustic rainfall rate algorithm is therefore more correctly a measure of the rainfall rate from large raindrops. Fortunately, the rainfall rate from large raindrops is highly correlated with the total rainfall rate, making acoustic monitoring of underwater sound an effective measure of rainfall rate in oceanic regions.

Orographic rainfall over low hills and associated corrections to radar measurements

The horizontal and vertical distribution of orographic rainfall over ridges of height approximately 150 m was resolved by a dense rain gauge network and rainfall radar. Orographic enhancements of up to about 2 mm h −1 were found in conditions typical of wintertime warm fronts and warm sectors in the UK. The enhancement magnitude (averaged over a few hours) was found to depend partly upon the surface wind and the average background rainfall rate. The growth of orographic precipitation was observed in cloud layers of varying depth; in one case, significant enhancement occurred within cloud with a total depth of no more than 1500 m. Doubt is cast upon the applicability of some existing models of the seeder-feeder mechanism to cases such as this. In this and other respects, our conclusions differ from earlier findings. The observations highlight the difficulty in estimation of surface rainfall rate by radar if the beam is several hundred metres or more above the surface. If no attempt is made to correct for precipitation growth beneath the beam, serious underestimation of rainfall is likely, even in lowland regions with hills of about 100 m height.