1. Introduction
There is growing evidence that the statistical characteristics of precipitation are changing at several places globally (Brunetti et al. 2001; Goswami et al. 2006; Trenberth 2011; Wang and Ding 2006; Yao et al. 2012). To understand the extent of these changes, accurate and timely knowledge of the space–time variability of precipitation is essential. Rain gauges provide direct and accurate point measurement of precipitation over land and are often the most used and trusted source of information for hydrological studies. However, one of the major shortcomings of gauge measurements is that they do not provide complete areal coverage and are often sparse or nonexistent in remote or politically unstable regions of the world. In addition, it is well known that gauge measurements are also prone to several sources of systematic and random error (Sevruk 1985; Kuligowski 1997). Given the importance of gauge-based precipitation estimates as “ground truth” for other products, significant progress has been made to develop and construct gridded global and regional gauge-based precipitation analyses.
Satellite-derived precipitation products have been developed that fill some of the data gaps by providing more spatially homogeneous and temporally complete coverage over large areas of ocean and land (Yilmaz et al. 2005; Dinku et al. 2007; Kucera et al. 2013). However, the accuracy of these products is limited, largely because they are derived from other observables (i.e., cloud-top reflectance or thermal radiance; Richards and Arkin 1981; Petty and Krajewski 1996). Satellite products can also suffer from various discontinuities between different platforms and instruments and are not available before the 1970s. Recently, various “merged” satellite and gauge analyses have been assembled that maximize (and minimize) the relative benefits (and shortcomings) of each data type (New et al. 2001).
Spatially homogeneous and temporally continuous precipitation estimates are also available from the reanalysis systems that use advanced data assimilation techniques to merge estimates from global circulation models with observations. Reanalysis products are widely used in climate research, even though the products are known to have biases, particularly in the estimation of precipitation (Ma et al. 2009; Lorenz and Kunstmann 2012).
In spite of their shortcomings, satellite-derived and reanalysis precipitation products offer an exciting opportunity to better understand the characteristics and variability of precipitation throughout the globe. However, before this potential can be fully realized, we need to validate these datasets and gain a better understanding of their inherent biases (Bosilovich et al. 2013; Turk et al. 2008). Intercomparisons of satellite-derived datasets and reanalyses with gauge observations have been used to assess the reliability of precipitation products at individual sites and for specific regions (Bosilovich et al. 2008; Ebert et al. 2007; Joshi et al. 2012; Kidd et al. 2012; Ma et al. 2009; Palazzi et al. 2013; Rahman et al. 2009; Shah and Mishra 2014; Shen et al. 2010). They showed that the performance of reanalyses and satellite-based precipitation estimates vary for different regions and for different precipitation regimes. In general, reanalyses showed greatest skill during winter seasons while satellite estimates were best over wet regions and for warm seasons.
Previous studies of precipitation over the Indian subcontinent have focused on the genesis, dynamics, diagnostics, and predictability of precipitation and have put the most emphasis on the Indian summer monsoon rainfall (ISMR; Krishnamurthy and Shukla 2000; Goswami and Ajaya Mohan 2001; Mishra et al. 2012; Turner and Annamalai 2012). Far less effort has focused on pre- and postmonsoon precipitation in spite of its importance to water resources, agriculture, and livelihoods in the northwestern/Himalayan and southern peninsular regions of the subcontinent (Immerzeel et al. 2009; Bookhagen and Burbank 2010). The magnitude and mechanistic explanation of precipitation and its variability during these seasons and/or regions remains unclear and poorly investigated (Kripalani and Kumar 2004; Sen Roy 2006; Rajeevan et al. 2012; Kar and Rana 2013).
In this study, we investigate the seasonal (winter, premonsoon, monsoon, and postmonsoon) precipitation climate and variability over the Indian subcontinent, using seven recent, high-quality, well-documented precipitation products from three different sources: gridded station data, satellite-derived data, and reanalyses. The study employs several verification methods to determine consistency and reliability associated with each precipitation product, in terms of availability, resolution, accuracy, retrieval technique, and quality control.
The main goals of this study are to 1) provide an intercomparison of the various precipitation datasets over our study region, 2) provide insights into data reliability and usability for a region that contains subregions that are not well provisioned by rain gauges, and 3) examine precipitation and its variability for all seasons throughout the year.
The results are mainly presented in the form of maps that highlight the key characteristics and regional differences associated with particular precipitation products. The paper is organized as follows. After the introduction in section 1, section 2 describes the datasets used, with a brief description of methodology and choice of reference dataset; and section 3 outlines the results and discussion, followed by the summary and conclusions in section 4.
2. Precipitation datasets
The main characteristics of the datasets used in this study are shown in Table 1. The basic selection criteria are based on spatial coverage over the study region, spatial resolution, and temporal availability.
Asian Precipitation–Highly-Resolved Observational Data Integration Towards Evaluation of the Water Resources (APHRODITE) was developed by a consortium of the Research Institute for Humanity and Nature (RIHN), Japan, and the Meteorological Research Institute of the Japan Meteorological Agency (MRI-JMA). APHRODITE is made up of high-quality daily precipitation products of varying resolution (0.25°, 0.5°) for several Asian subregions. Here, we use the latest version of daily precipitation data (APHRO_V1101) at 0.5° latitude–longitude resolution for the Asian monsoon domain (15°S–55°N, 60°–150°E) during the period 1951–2007 (available from www.chikyu.ac.jp/precip/products/). The product does not discriminate between rain and snow but incorporates an improved quality-control method and orographic correction of precipitation. Details on the gridding procedure and reliability of APHRODITE daily precipitation data can be found in Hamada et al. (2011) and Yatagai et al. (2009, 2012).
Climate Prediction Center unified (CPC-uni) is a global gauge-based daily precipitation product from the National Oceanic and Atmospheric Administration (NOAA) Climate Prediction Center (CPC). Gauge reports from over 30 000 stations are collected from multiple sources, including Global Telecommunications System (GTS), Cooperative Observer (COOP) network, and other national and international agencies, and quality control is performed. CPC-uni uses optimal interpolation (OI) with orographic consideration to represent the area-averaged values of precipitation over the grid boxes (Chen et al. 2008). Here, we used the CPC-uni, version 1.0 (v1.0), global land data at a 0.5° latitude–longitude spatial resolution available from 1979 to the present.
The Global Precipitation Climatology Project (GPCP), sponsored by the World Climate Research Programme and Global Energy and Water Cycle Experiment, provides global precipitation products based on satellite and gauge information at daily (Huffman et al. 2001), pentad (Xie et al. 2003), and monthly (Adler et al. 2003) time scales. The GPCP 1-Degree Daily (1DD), version 1.2 (v1.2), dataset covers the period from October 1996 up to the present and is a companion to the GPCP, version 2.2, satellite–gauge (SG) combination. Information on individual components used as an input in the GPCP-1DD precipitation product can be obtained online (http://precip.gsfc.nasa.gov/gpcp_daily_comb.html).
The Tropical Rainfall Measuring Mission (TRMM) is a joint mission between the Japan Aerospace Exploration Agency (JAXA) and the U.S. National Aeronautics and Space Administration (NASA) designed to monitor precipitation and its variability within the tropics. The TRMM Multisatellite Precipitation Analysis (TMPA) algorithm merges a variety of satellite-based observations and ground observations to yield high spatiotemporal resolution and quasi-global quantitative precipitation estimation (QPE) products. The 3-hourly TMPA precipitation estimates are available as a real-time version (3B42-RT) and a gauge-adjusted post-real-time research version (3B42). There are two generations of 3B42 products: version 6 (3B42-V6), which is available to June 2012; and version 7 (3B42-V7), which was retrospectively processed using the updated algorithm back to 1998 and released in May 2012 and is ongoing. The newly released 3B42-V7 incorporates several important changes over 3B42-V6, including use of Global Precipitation Climatology Centre (GPCC) analyses and additional satellite data (Huffman and Bolvin 2014). For this study, we use both the research versions (hereafter referred to as 3B42-V6 and 3B42-V7) available through the NASA interface (mirador.gsfc.nasa.gov). Both datasets have a spatial (temporal) resolution of 0.25° × 0.25° (3 h) over latitudes 50°N–50°S (Huffman et al. 2007, 2011).
The Climate Forecast System Reanalysis (CFSR) is a third-generation reanalysis product developed at the National Centers for Environmental Prediction (NCEP; Saha et al. 2010). The advantages of CFSR relative to previous NCEP reanalyses include (i) higher horizontal (~38 km, T382) and vertical (64 sigma-pressure hybrid levels) resolution, (ii) the guess forecast is generated from a coupled atmosphere–land–ocean–ice system, and (iii) historical satellite radiance measurements are assimilated. In addition, the CFSR is forced with observed estimates of evolving greenhouse gas concentrations, aerosols, and solar variations (http://cfs.ncep.noaa.gov/cfsr/). This study uses daily precipitation at ~0.313° Gaussian latitude–longitude, available for the period of 1979–2010 from the National Center for Atmospheric Research (NCAR).
The European Centre for Medium-Range Weather Forecasts (ECMWF) interim reanalysis (ERA-Interim) is the most recent global atmospheric reanalysis produced by ECMWF covering the period from 1979 onward (Dee et al. 2011). ERA-Interim uses four-dimensional variational data assimilation (4D-Var), a revised humidity analysis, variational bias correction for satellite data, and other improvements in data handling. ERA-Interim precipitation estimates are produced by the forecast model based on temperature and humidity information derived from assimilated observations. The dataset is freely available at a global daily resolution of ~1.5° regular Gaussian grid.
Main characteristics of the precipitation products examined in this study.
a. Data processing and method
No standard format or grid resolution exists among the various precipitation products (Table 1). To facilitate direct comparison of precipitation datasets, a common 0.5° × 0.5° regular latitude–longitude grid was chosen and bilinear interpolation was used. CFSR and ERA-Interim were extracted from the Gaussian grid gridded binary (GRIB) format and remapped to the (0.5° × 0.5°) regular grid. Seasonal averages were calculated from daily precipitation values and extracted for the region of interest, the Indian subcontinent (5°–38°N, 60°–100°E; Fig. 1). Only land points are considered here and the seasons are defined as winter [December–February (DJF)], premonsoon [March–May (MAM)], monsoon [June–September (JJAS)], and postmonsoon [October–November (ON)]. These seasonal designations best segment the large-scale annual variations of the major components of the subcontinent’s climate, such as surface wind/surface pressure, temperature, and precipitation (Shea and Sontakke 1995). A common overlap period of 10 years (1997/98–2006/07) was chosen to enable a consistent comparison.
Location and topography (m) of the Indian subcontinent.
Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0106.1
Several statistical methods are employed to analyze the precipitation products compared to the reference APHRODITE data. Maps are generated for seasonal and monthly mean climatology, standard deviation (std dev), and anomaly correlation coefficients (ACC). Percentage precipitation difference (PPD), defined as 100[(product − reference data)/reference data], and percentage root-mean-square error (PRMSE), defined as 100(RMSE/reference data), are computed to measure the biases in all examined products, relative to the seasonal mean of the reference dataset. To examine the spatial–temporal variability of precipitation in each dataset, empirical orthogonal function (EOF) analysis is applied. This technique seeks to reduce the dimensionality of a dataset by finding a new set of variables that capture most of the observed variance in the data through a linear combination of original variables (Wilks 1995). The derived pattern of variability in space is known as the EOF, and in time it is known as principal component (PC). EOF analysis has been widely used in the field of climate research for many years (Ding and Wang 2005; Mishra et al. 2012; Zhang et al. 2013).
b. Why APHRODITE as the reference?
In the present study, we have considered APHRODITE as our primary surface reference (i.e., benchmark) dataset. Yatagai et al. (2008, 2009) recommended APHRODITE as a reliable dataset for studying rainfall variations over Asia and suggested that it can be considered as observations and ground truth around the Himalayas, the Middle East, and central Asian regions. Rajeevan and Bhate (2009), in a study over the Indian region, compared the India Meteorological Department (IMD) rain gauge data (comprising 6000 stations) against the APHRODITE_V0804 rainfall product (~2000 stations) and found that APHRODITE data could capture the large-scale features of the ISMR, exhibiting strong correlations (>0.6) with IMD data over most of India. For monitoring large-scale precipitation variations over East Asia, Sohn et al. (2012) quantified the reliability of six precipitation datasets over the region considering APHRODITE as the ground truth. Andermann et al. (2011) compared gridded precipitation products along the complex relief zone of the Himalayan front and concluded that for all seasons the APHRODITE dataset gives the best precipitation estimates with minimal difference from independent ground observations. However, the authors noted that the lack of gauge stations at higher elevations of the study region limits the accuracy of APHRODITE. Recently, Prakash et al. (2015) performed a seasonal (JJAS) intercomparison of six observational rainfall datasets against IMD gridded data and found that APHRODITE and GPCC are the two best performers in terms of statistical skill scores, with APHRODITE’s root-mean-square error (39.31%) being lower than that of GPCC (42.68%) for all-India JJAS rainfall.
These studies, together with the validation work carried out by Yatagai and Xie (2006) and Yatagai et al. (2009), give us a sound basis to consider APHRODITE as a reliable reference dataset for studying precipitation variations over the Indian subcontinent and other regions, where the dataset contains a dense network of rain gauges. Figure 2 shows the distribution map of rain gauge stations used by APHRODITE for the year 1998 (Yatagai et al. 2009). As seen in Fig. 2, the dataset benefits from a dense network of rain gauges over a large part of India and Nepal, giving greater confidence for diagnostic studies and validation work. However, over the northwestern region of the subcontinent (including northwestern India, Pakistan, and Afghanistan), the eastern part of peninsular India, and the Tibetan Plateau, the distribution of gauge stations is sparse, which may limit the accuracy of observations and might introduce uncertainties in gauge-based precipitation products.
Distribution of rain gauge stations used by APHRODITE for the year 1998: GTS stations (blue), precompiled datasets (black), and data individually collected by the APHRODITE project team (red).
Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0106.1
3. Results and discussion
a. Seasonal to interannual variability
The Indian subcontinent exhibits a wide range of seasonal precipitation regimes across different regions. To visualize this distribution of total annual precipitation, daily precipitation values for the time period from December 1997 to December 2007 are segmented into four seasons (DJF, MAM, JJAS, and ON) and spatial fields of seasonal precipitation distribution are generated for all datasets (Fig. 3). The most pronounced spatial precipitation pattern represents the southwest monsoon (JJAS) rainfall that contributes ~70%–75% of the total annual precipitation across the entire domain. Rainfall during the southwest monsoon is a result of both the seasonally changing thermal contrast between land and ocean and the seasonal migration of the intertropical convergence zone (ITCZ). The highest climatological mean for JJAS rainfall occurs over the western coast of peninsular India and over the northeastern region; both are associated with orographic lifting and intense convection (Pattanaik and Rajeevan 2010).
Spatial distribution of seasonal mean precipitation (mm day−1) over the Indian subcontinent for each precipitation product for the period 1997/98–2006/07.
Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0106.1
The JJAS rainfall gradually decreases northwestward from east-central India, with areas along the southeastern coast of peninsular India being comparatively dry. It is during the postmonsoon season that large parts of the southeastern coast of peninsular India and neighboring Sri Lanka receive a significant amount of rain due to the retreating southwest monsoonal winds, also known as the northeast monsoon. With the retreat of the southwest monsoon, surface pressure and the upper wind circulation patterns change rapidly and a trough of low pressure is established over the southern Bay of Bengal, resulting in a northeasterly wind flow across the subcontinent (Srinivasan and Ramamurthy 1973). During this time, the northern–northwestern part of the subcontinent is generally dry but receives a significant amount of precipitation during winter (DJF) and premonsoon (MAM) months because of the occurrence of midlatitude disturbances and premonsoon thunderstorms (TSs), respectively. Winter precipitation is usually limited in spatial extent and total amount and is largely confined to the northwestern region of the Indian subcontinent and follows a general decline from west to east (Sen Roy 2009). An interesting feature observed during the premonsoon months is the shift in rainfall activity to the northeastern region of the subcontinent as a result of the TS activity before the advancement of the southwest monsoon over the region (Mahanta et al. 2013).
Figure 3 shows that all products are capable of coherently reproducing the key features of the seasonal and regional precipitation distribution reasonably well and are roughly in agreement with the APHRODITE dataset. However, significant regional differences can be observed in all precipitation products, especially along the topographically induced high rainfall regional features (e.g., over northern Pakistan, the foothills of the Himalayas, the extreme northeast, and along the Western Ghats). Comparing APHRODITE with other products, we find that CPC-uni, GPCP, and 3B42-V6 capture the seasonal distribution pattern of DJF and MAM precipitation, but the datasets fail to capture the maxima of orographic precipitation over the complex topography of northwestern India. In addition, the products slightly underestimate (overestimate) the rainfall amounts over the windward (leeward) side of the western coast of India during the JJAS season.
In mountainous regions, 3B42-V7 exhibits slightly higher precipitation and hence performs better than 3B42-V6, CPC-uni, and GPCP. According to Huffman and Bolvin (2014), the largest difference between 3B42-V7 and 3B42-V6 occurs in mountainous regions owing to the major changes in 3B42-V7’s Precipitation Radar algorithm, which uses a higher-resolution elevation map based on Shuttle Radar Topography Mission data with 30-arc-s spacing (SRTM30) over India, Tibet, and South America (TRMM Precipitation Radar Team 2011). The newer product (3B42-V7) displays the rain-shadow region of the Western Ghats reasonably well but fails to capture the distribution pattern of postmonsoon rainfall along the eastern coast of peninsular India. Moreover, 3B42-V7 exhibits slightly higher precipitation over the core monsoon zone and the northeastern region of the subcontinent, including Myanmar. Reasons for overestimation in 3B42-V7 could be bias in the full usage of GPCC gauge data, which, according to Prakash et al. (2015), overestimates (~5 mm day−1) the monsoon rainfall over northeastern India and along the Western Ghats compared to IMD gauge data, or because of the occurrence of heavy convective rainfall, which may cause signal attenuation and might lead to overestimation of near-surface rainfall (Amitai et al. 2004).
In contrast to CPC-uni and the satellite-derived precipitation products (GPCP, 3B42-V6, and 3B42-V7), the two reanalysis products (CFSR and ERA-Interim) appear to reproduce high seasonal precipitation values over regions of pronounced precipitation activity. For example, high precipitation totals are observed in the northwestern and northeastern regions of the subcontinent for DJF and MAM. In addition, regions of spuriously high rainfall are also seen during JJAS, particularly over the northeastern region and the core monsoon zone. Monsoon precipitation in these regions is mainly associated with the presence of a moist convective regime related to the monsoon flow and also to the propagation of low pressure systems from the Bay of Bengal along the core monsoon zone (Rajeevan et al. 2010). Therefore, overestimation of precipitation in these regions, especially by reanalysis datasets, could be associated with overestimation of moisture fields and hence precipitable water (Trenberth et al. 2011; Shah and Mishra. 2014). CFSR tends to overestimate seasonal precipitation in the southern part of peninsular India during MAM, JJAS, and ON and to underestimate JJAS precipitation over the semiarid region of northwestern India (including Pakistan), consistent with Shah and Mishra (2014).
The reanalyses illustrate the strong influence of orography on precipitation. While ERA-Interim shows a smooth spatial distribution of precipitation, CFSR reproduces finer-scale regional patterns, which may be due to the improved representation of the orographically influenced precipitation and the high spatial resolution of this dataset (Saha et al. 2010). A close comparison of precipitation products illustrates that 3B42-V6 and CPC-uni display the highest degree of spatial agreement to APHRODITE’s (JJAS and ON) seasonal climatology. For the entire subcontinent, the seasonal JJAS (ON) domain-averaged mean is 4.29 (1.31) for APHRODITE, 3.98 (1.22) for CPC-uni, 5.2 (1.84) for GPCP, 4.35 (1.36) for 3B42-V6, 5.9 (1.11) for 3B42-V7, 7.1 (2.03) for CFSR, and 5.8 (1.89) mm day−1 for ERA-Interim.
The mean monthly precipitation cycles over the region for the period 1997/98–2006/07 are compared in Fig. 4. All products capture the unimodal distribution of the annual cycle fairly well, albeit with different amplitude. With respect to the APHRODITE, CPC-uni (GPCP) tends to slightly underestimate (overestimate) the monthly precipitation amount, while CFSR and ERA-Interim significantly overestimate mean monthly precipitation from May to December. The 3B42-V7 closely agrees with APHRODITE’s averages for the winter and premonsoon months, but overestimates (slightly underestimates) the JJAS (ON) precipitation amounts over the core monsoon zone (eastern coast) of India. The results seen here are a clear manifestation of the spatial climatology patterns depicted in Fig. 3.
Mean monthly precipitation cycle (mm day−1) averaged over the Indian subcontinent for each precipitation product for the period 1997/98–2006/07.
Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0106.1
To represent the interannual variability of precipitation at each grid point, we generate interannual std dev plots for each season (Fig. 5). All datasets reveal large std dev values over regions that receive large amounts of seasonal precipitation. For example, the western coast of peninsular India and the northeastern region exhibit large std dev values during JJAS, while southeastern peninsular India displays high interannual variability during ON. Similarly for DJF and MAM, comparatively high (>5 mm day−1) std dev values appear over the northwestern region and in some parts of northeastern and southern peninsular India. In comparison to APHRODITE, the overall magnitude of rainfall variability is fairly well reproduced by all datasets for all seasons and/or regions, with the exception of CFSR. During JJAS, CFSR displays large (>18 mm day−1) interannual variability over north-central India and northeastern regions of the subcontinent, where CFSR produces excess precipitation totals (Fig. 3). Misra et al. (2012) commented that CFSR displays the strongest interannual variation of precipitable water over central India (with nearly 11% difference between wet and dry years), which they associated with the anomalous oceanic sources of moisture from the adjacent Bay of Bengal and Arabian Sea. In the JJAS (std dev) pattern, an important feature in ERA-Interim is the region of suppressed (<9 mm day−1) rainfall variability along the western coast of India and comparatively high (16 mm day−1) variability over the northeastern region.
Std dev of seasonal mean precipitation (mm day−1) over the Indian subcontinent for each precipitation product for the period 1997/98–2006/07.
Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0106.1
The results obtained here are consistent with the finding of Shah and Mishra (2014), who reported that both CFSR and ERA-Interim overestimated monsoon rainfall variability over eastern regions compared to IMD gauge observations. For the monsoon season, the domain-averaged std dev is APHRODITE = 10.4, CPC-uni = 10.2, GPCP = 9.6, 3B42-V6 = 11.2, 3B42-V7 = 11.4, CFSR = 17.3, and ERA-Interim = 11.3 mm day−1. In comparison to GPCP, the relative superiority of CPC-uni, 3B42-V6, and 3B42-V7 is evident, as these products show close agreement to APHRODITE’s precipitation variability and can capture the pattern of large std dev values over the Western Ghats (eastern coast) of India during JJAS (ON). Rahman et al. (2009) and Joshi et al. (2012) also reported that in comparison to IMD gauge data, 3B42-V6 performs better than GPCP for the monsoon period over the Indian region. Among the reanalysis datasets, the performance of ERA-Interim looks somewhat better than CFSR in terms of both domain-averaged mean and std dev for the JJAS period. For DJF and MAM, CPC-uni and GPCP do not provide any useful information with regard to the interannual variability of seasonal precipitation over the northwestern region of the Indian subcontinent and along the foothills of the Himalayas. Despite the similar std dev distribution pattern for both 3B42-V6 and 3B42-V7, differences are notable over mountainous regions, where 3B42-V7 more successfully captures the interannual variability of the DJF and MAM precipitation as seen in APHRODITE. The domain-averaged std dev values for DJF are APHRODITE = 3.4, CPC-uni = 2.9, GPCP = 3.1, 3B42-V6 = 3.1, 3B42-V7 = 3.6, CFSR = 4.3, and ERA-Interim = 4.3 mm day−1.
b. Error assessment
To explore the skill of the precipitation products on a regional basis, a relative bias or PPD (relative to APHRODITE) is computed on a seasonal scale. Figure 6 indicates that CPC-uni exhibits a general tendency to underestimate seasonal precipitation along the foothills of the Himalayas, the Tibetan Plateau, and Myanmar in all seasons. The poor performance of the CPC-uni indicates that the total number of active rain gauges and their spatial distribution used to construct the CPC-uni dataset is not enough to capture the seasonal precipitation pattern and interannual variability over the above-mentioned regions. According to Gebregiorgis and Hossain (2015), APHRODITE uses an extensive network of gauges from various partner organizations, which is more than CPC-uni.
PPD with respect to APHRODITE seasonal mean precipitation over the Indian subcontinent for the period 1997/98–2006/07.
Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0106.1
Large PPDs appear over the data-sparse and topographically complex Tibetan Plateau and the Hindu Kush–Karakoram mountain ranges in the northwestern region of the subcontinent. The domain-averaged PPDs are maximum (minimum) for the winter (monsoon) season with values of 13% (4%), 142% (53%), 43% (25%), 72% (38%), 243% (41%), and 234% (84%) for CPC-uni, GPCP, 3B42-V6, 3B42-V7, CFSR, and ERA-Interim, respectively. Similarly for MAM (ON), the domain-averaged PPDs are 2% (4%), 63% (68%), 33% (28%), 40% (45%), 79% (103%), and 123% (114%), respectively. The large positive bias over the above-mentioned regions appears to be consistent in the reanalysis datasets throughout the year. The large positive bias in CFSR and ERA-Interim could be related to 1) the inability of the models to resolve orographic precipitation in complex terrain with large elevation changes (Ma et al. 2009), 2) differences in the land–atmosphere interaction and land surface model schemes (Misra et al. 2012), and 3) the fact that reanalysis products estimate total precipitation (rainfall plus snow; Palazzi et al. 2013). However, for the satellite-derived precipitation products, the large bias seems to gradually decrease from the cold winter months (DJF) to the warm summer season (JJAS). This result closely agrees with the conclusions reported by Ebert et al. (2007) that satellite algorithms perform best during summers, when rainfall is more convective, than in winters when it is more synoptically forced. Apparently, the occurrence of comparatively smaller positive bias for the satellite-derived precipitation products could be related to the difficulty of in situ station data, satellite estimates, and their combinations in detecting the snow component of precipitation (Rasmussen et al. 2012).
All datasets (except for the gauge-only CPC-uni) display distinctly large winter PPDs over the northern flank of the subcontinent and the Tibetan Plateau. This implies an additional source of uncertainty (i.e., gauge measurements) might be contributing toward the existing errors in the reanalysis estimates and satellite retrievals. According to Lorenz and Kunstmann (2012), gauge undercatch error for solid precipitation, which is especially large in high latitudes or mountainous regions during the winter season, can lead to an underestimation of the true precipitation of up to 50%. Moreover, sampling errors between 15% and 100% could be expected for sparsely gauged regions with less than three gauges per 2.5° × 2.5° grid cell (Rudolf and Rubel 2005). Therefore, depending on the environmental conditions (season) and geographic location, it is likely that the uncertainties and inhomogeneities in observational/reference data are artificially inflating the existing biases in all datasets. Furthermore, a comparison of the spatial and domain-averaged PPDs indicates that 3B42-V6 and 3B42-V7 outperform GPCP in all seasons. This can be associated first to the fact that, in regions with low gauge density, the GPCP analysis is influenced more by the gauges (Gruber et al. 2000), while 3B42 estimates are influenced more by the satellite measurements, and second, to the coarser resolution of GPCP, which spreads any errors over larger geographical areas. CFSR displays large negative PPDs over Afghanistan and Pakistan and some parts of northwestern India during JJAS. Misra et al. (2012) found that during the monsoon season the cross-equatorial and the southwesterly flow of monsoonal (950 hPa) winds is weak in CFSR, which, according to our understanding, could result in the weakening of the monsoon flow and hence less advection of moisture to the far northwestern regions (including Pakistan), as seen in Fig. 3.
Spatial patterns of ACC are generated for the period from 1997/98–2006/07 in Fig. 7 to analyze how the examined products relate to the seasonal anomalies in APHRODITE data at each grid point. Results indicate that for a season, the ACCs are significantly high (>0.632 at 95% significance level) over regions exhibiting less interannual variability and vice versa. Reasonably high (low) ACCs are observed during ON (JJAS) over most of the regions and for all products. In comparison to other precipitation products, CPC-uni displays relatively high and significant ACC values (>0.65) along the western (eastern) coast of peninsular India during JJAS (ON), suggesting that the two gauge products have similar anomaly pattern. The domain-averaged ACC for CPC-uni and APHRODITE is found to be DJF = 0.36, MAM = 0.57, JJAS = 0.47, and ON = 0.63. In addition, 3B42-V6 and 3B42-V7 also show a close agreement to APHRODITE during MAM, JJAS, and ON, indicating a comparably good anomaly correlation in regions exhibiting high precipitation variability (e.g., Western Ghats and northeastern region including Myanmar). The domain-averaged ACCs for 3B42-V6 (3B42-V7) during DJF, MAM, JJAS, and ON are found to be 0.21 (0.21), 0.54 (0.57), 0.46 (0.48) and 0.67 (0.68), respectively. Unlike other precipitation products, 3B42-V6 and 3B42-V7 exhibit the lowest domain-averaged ACCs (i.e., 0.21) and display a different anomaly correlation pattern during the DJF season. The diagonal correlation pattern with positive–negative–positive ACC values is similar to the high–low–high regional precipitation seasonality pattern generated in Fig. 3. However, the negative ACC values in 3B42-V6 and 3B42-V7 are of concern, indicating in some cases the reversed anomalies of winter precipitation over the Indian subcontinent.
ACC of seasonal mean precipitation over the Indian subcontinent compared to APHRODITE for the period 1997/98–2006/07.
Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0106.1
In general, it appears that all individual precipitation products have a relatively low and seasonally variable ACC spatial pattern over the northwestern region of subcontinent, the entire Himalayan belt, the Tibetan Plateau, and Myanmar. The two reanalyses products more or less represent similar characteristic features in their ACC patterns and produce poor correlations during MAM (CFSR = 0.35, ERA-Interim = 0.42) and JJAS (CFSR = 0.29, ERA-Interim = 0.30) as compared to other precipitation products. This result perhaps can be attributed entirely to the strong overestimation of precipitation (Figs. 3, 4) and high interannual variability (Fig. 5) over the core monsoon zone and the northeastern region of the Indian subcontinent, which largely reduces the skill of the reanalyses products over these regions.
Figure 8 shows the seasonal PRMSE with respect to the climatological mean of the APHRODITE dataset. The PRMSE ranges from zero to infinity, with lower values indicating closer agreement with APHRODITE. The maximum (minimum) error/uncertainty in precipitation estimates exists during the winter (monsoon) period. More or less consistent with the PPD results, the spatial comparison of PRMSE reveals that, regardless of the season, PRMSEs and PPDs are large and distinct over the topographically complex and data-sparse northwestern region of the Indian subcontinent, Myanmar, and the Tibetan Plateau. Conversely, the spatial spread of PRMSE is homogeneous and generally smaller over low-elevation plain areas where moderate to dense gauge stations are available. For the subcontinent as a whole, the domain-averaged PRMSEs for the winter season are much greater in the reanalysis datasets (CFSR = 335%, ERA-Interim = 311%), than for CPC (104%), 3B42-V6 (143%), 3B42-V7 (161%), and GPCP (209%). Similarly for the JJAS season, the PRMSEs for CPC-uni, GPCP, 3B42-V6, 3B42-V7, CFSR, and ERA-Interim are found to be 61%, 78%, 62%, 65%, 100%, and 171%, respectively. For the pre- and postmonsoon seasons, the PRMSE is <90% for CPC-uni, 3B42-V6, and 3B42-V7; and >150% for CFSR, ERA-Interim, and GPCP. Apparently for the monsoon season, the PRMSE spatial spread and domain-averaged score is a minimum in CPC-uni over the Indian region, while among the reanalysis products, ERA-Interim appears to be slightly better than CFSR in terms of the spatial spread of PRMSEs over the Indian region alone. Additionally, 3B42-V6 and 3B42-V7 seem to outperform GPCP again, with regard to the spatial representation of comparatively smaller errors over the Western Ghats, the foothills of the Himalayas, and Myanmar during the monsoon season. The extreme spatial PRMSEs observed during the “cold” winter season mentioned above illustrate important limitations in all gridded precipitation products examined here (including the APHRODITE dataset). Based on the results obtained in Figs. 6 and 8, together with the well-known fact that gauge observations are subject to several sources of systematic and random errors, it is likely that an additional source of error could be associated with the gauge input data (especially when and where gauge stations are sparse and/or snow makes up a large portion of the total precipitation).
PRMSE with respect to APHRODITE seasonal mean precipitation over the Indian subcontinent for the period 1997/98–2006/07.
Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0106.1
c. EOF analysis
To explore the structure of large-scale coherence within the precipitation datasets, we use EOF analysis on the standardized seasonal precipitation anomalies. Results are shown in Fig. 9. Here we have focused on the first leading mode (EOF1), which accounts for the largest percentage of the total precipitation variance, represented at the bottom left of each spatial plot. While interpreting the EOF patterns, it must be kept in mind that the sign of the patterns is arbitrary (the temporal mean of their PC time series is zero). The spatial pattern of the leading EOF during DJF reveals a large-scale pattern of negative loading over most of the Indian subcontinent with some positive loading over the southern tip of peninsular India, north of the Tibetan Plateau, and over some parts of Myanmar. For APHRODITE, the first winter EOF explains 53% of the total variance, which is comparable to that of GPCP (50%). The loading patterns of EOF1 are very much similar for 3B42-V6 and 3B42-V7 in all seasons, albeit with differences in the percentage of variance explained (always higher for 3B42-V7). For DJF, the EOF pattern obtained from CFSR and ERA-Interim resembles APHRODITE’s spatial pattern more closely than the one obtained from GPCP, 3B42-V6, and 3B42-V7.
Leading EOF of standardized seasonal mean precipitation over the Indian subcontinent for the period 1997/98–2006/07.
Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0106.1
Kar and Rana (2013) used EOF analysis to investigate the impact of the Asian jet on interannual variability of winter precipitation over northwestern India. The authors found that the PC1 and PC2 time series corresponding to EOF1 and EOF2 of DJF zonal winds at 200 hPa represented a pattern similar to El Niño–Southern Oscillation (ENSO) and the North Atlantic Oscillation (NAO)–Arctic Oscillation (AO), respectively, with a statistically significant correlation between the PC time series and winter precipitation. Syed et al. (2009) also found that both ENSO and NAO have a significant influence on winter precipitation over southwest-central Asia (including northern Pakistan, Afghanistan, and Tajikistan), where winter precipitation increases (decreases) during the warm (cold) ENSO phase and positive (negative) NAO phase. The time coefficient of our leading EOF, that is, the PC1 of the standardized DJF precipitation anomalies, is shown in Fig. 10a. Consistent with the above findings, we found that the DJF-PC1 time series obtained for all precipitation products corresponds well with the DJF–Niño-3.4 index obtained from the CPC (www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml) for the period 1997/98–2006/07. The correlation between the DJF-PC1 time series and the Niño-3.4 index is found to be statistically significant and greater than 0.7 for all datasets used here.
PC time series for the leading EOFs of standardized seasonal mean precipitation for (a) winter, (b) premonsoon, (c) monsoon, and (d) postmonsoon over the Indian subcontinent for each precipitation product for the period 1997/98–2006/07.
Citation: Journal of Hydrometeorology 16, 2; 10.1175/JHM-D-14-0106.1
In MAM, we find that the small area of positive loading over the northwestern region of the subcontinent seen in DJF is persistent and more prominent during MAM for all precipitation products. The loading represents a seesaw-like pattern of positive (negative) loading over the northwestern region (the rest of the subcontinent), which shows similarities with the seasonal mean precipitation distribution pattern over the subcontinent (Fig. 3). The percentage variance explained by the precipitation products is either comparable to or higher than that of APHRODITE (37%). However, the structure of loading pattern is slightly different for the reanalysis datasets, with a broader region of positive loading over the northeastern domain, which is not very prominent in any other dataset. Therefore, during MAM the structure of EOF1 from APHRODITE most closely resembles those generated from CPC-uni and the satellite-derived products. The corresponding MAM-PC1 time series (Fig. 10b) is similar to the DJF-PC1 sequence and exhibits the interannual variability present in the premonsoon domain-averaged precipitation time series (figure not shown). All datasets capture the 3-yr (1998–2001) extended winter and premonsoon drought that lasted over Afghanistan, Iran, and Pakistan. According to Barlow et al. (2002), the prolonged drought was linked to the cold phase of ENSO (La Niña) and the unusually warm pool region in the western Pacific Ocean.
The JJAS-EOF1 patterns (Fig. 9) are characterized by uniform variability over the entire Indian subcontinent. In general, for all precipitation products, the JJAS pattern reveals positive loading over a large part of the subcontinent, with small areas of negative loading over central India and in the northeastern region of the subcontinent. Mishra et al. (2012) reported that the region, for which the JJAS monsoon rainfall oscillates in phase with the leading PC1 time series, includes a large part of the Indian region along with Pakistan, most of the Arabian Sea, and the Bay of Bengal. The percentage of variance explained by the leading EOFs for the satellite-derived products (GPCP = 31%, 3B42-V6 = 35%, and 3B42-V7 = 37%) are close to one another but higher than the variance obtained for the two reanalyses products (27%) and APHRODITE (26%) dataset. Surprisingly, the APHRODITE JJAS-EOF1 pattern differs in its spatial structure when compared to other products. The expected pattern [as reported by Krishnamurthy and Shukla (2000), Guhathakurta and Rajeevan 2008, Joshi et al. (2012), and Mishra et al. (2012)] is quite well captured by CPC-uni and the satellite-derived products, exhibiting a uniform structure of positive loading (0.02–0.08) over a large part of the Indian region and some pockets of negative loading over central and northeastern India. We further investigated the EOF pattern obtained for APHRODITE using a longer time series (1979–2007) and found that the leading EOF showed a close resemblance to the expected pattern. Therefore, it can be suggested that the observed difference in APHRODITE’s EOF1 pattern is likely due to the short time of analysis considered here. Most studies cited above show that the PC1 time series of JJAS rainfall is strongly correlated with the all-India-averaged rainfall index; which suggests that EOF1 is a good representation of the dominant mode of the JJAS precipitation variation (flood and drought years). When the standardized JJAS-PC1 series (Fig. 10c) are correlated with their respective seasonal precipitation mean time series, the maximum correlation is found for CPC-uni (0.98) and minimum for APHRODITE (0.38). Apparently for the APHRODITE dataset, the relatively poor correlation relates to the inability of the dataset to accurately capture the leading JJAS-EOF1 pattern. Additionally for GPCP, 3B42-V6, and 3B42-V7 the correlation is found to be >0.75, while for CFSR and ERA-Interim it is 0.68 and 0.5, respectively.
For the postmonsoon season, the spatial structure of EOF1 is similar to that in MAM but with a pronounced northwest–east and northwest–southeast gradient of positive–negative–positive loading over the subcontinent. The region of positive loading over southern peninsular India is relatively small for all products except for CFSR and ERA-Interim. Nair et al. (2013) used EOF analysis of eight global circulation models and found that during October–December the southern peninsular zone remains spatially correlated to the first EOF mode with positive loading over a large region. Kumar et al. (2007) found that only in recent years (1976–2000) has there been an increased relationship between ENSO and the northeast monsoon rainfall occurring over the southern peninsular region of India. Here, the PC1 time series (Fig. 10d) is found to be significantly (95%) correlated with the corresponding seasonal precipitation time series over the entire subcontinent. However, we did not find a significant correlation between the ENSO index and the PC1 time series, which could possibly be related to the shorter time series of the examined products considered here or due to the problem of domain dependency of EOF solutions (Horel 1981; Richman 1986).
4. Summary and conclusions
The objective of this study was to perform an intercomparison of precipitation products over the Indian subcontinent and to provide useful insights on the reliability and usability of precipitation datasets, over regions that are not well provisioned by rain gauges. To this end, seven precipitation products (APHRODITE, CPC-uni, GPCP, 3B42-V6, 3B42-V7, CFSR, and ERA-Interim) were examined in regard to the seasonal (DJF, MAM, JJAS, and ON) precipitation climate and variability for the period 1997/98–2006/07.
a. General comments
All products performed reasonably well in capturing the main characteristic features of seasonal precipitation variability, particularly over low-elevation plain areas where APHRODITE contains a moderate to dense network of rain gauge stations. Referring to the gauge distribution map in Fig. 2, there are certain regions/areas where the rain gauge network is dense and perhaps more reliable to intercompare and ascertain the accuracy and biases in the proxy datasets. Hence, over these regions, the data user can expect to have greater confidence in the accuracy and certainty of the results presented here.
Depending on the environmental conditions (season) and geographic location (complex terrain), gauge measurements are subject to several sources of errors that largely limit the accuracy of gauge measurements. Over regions where the spatial coverage of gauge observation is poor, proxy datasets can provide a potential source of information, but their accuracy and reliability can be considered as examined only. Our results show that over the topographically complex and data-sparse regions, pronounced regional and seasonal differences exist between the examined products and the APHRODITE dataset. Based on our relative comparisons (PPD and PRMSE), it may be concluded, though with lesser confidence that in gauge-sparse regions, a potential source of uncertainty in all gauge-based precipitation products is likely associated with the uncertainties and inhomogeneities in gauge measurements itself.
b. Our interpretation of seasonal comparisons
For the monsoon season (JJAS), results indicate that the gauge-only (CPC-uni) and satellite-derived precipitation products (GPCP, 3B42-V6, and 3B42-V7) capture the JJAS rainfall variability better than the reanalyses (CFSR and ERA-Interim). The 3B42-V6 displayed a close agreement to APHRODITE, with regard to spatial and domain-averaged statistics for the entire Indian subcontinent. Additionally, 3B42-V7 and CPC-uni demonstrated a reasonable performance over the entire subcontinent and Indian region, respectively. Among the satellite-derived precipitation products, the two post-real-time research products of 3B42 (V6 and V7) outperformed GPCP not only for the monsoon period, but also throughout the year. On the contrary, CFSR and ERA-Interim exhibited a high positive bias over the northeastern region of the subcontinent and the northern flanks of the Tibetan Plateau.
Similar conclusions can be drawn for the postmonsoon season (ON), where reanalyses (especially CFSR) overestimate the seasonal precipitation totals over parts of peninsular India, the northeastern region of the subcontinent, and the Tibetan Plateau. The 3B42-V6 product displays the most reasonable precipitation total, closer to APHRODITE, while 3B42-V7 underestimates the seasonal precipitation amount in comparison to APHRODITE and other examined products. Results show that the gauge-based CPC-uni largely underestimates precipitation along the foothills of the Himalayas, the Tibetan Plateau, and Myanmar in all seasons. To improve the quality of the product, both in accuracy and reliability over these regions, use of additional rain gauge stations is highly desirable.
Likewise, for the premonsoon season (MAM), the satellite-derived precipitation products perform better than the reanalyses estimates. Relative to the APHRODITE dataset, CFSR and ERA-Interim consistently overestimate precipitation over the topographically complex and data-sparse Tibetan Plateau and the Hindu Kush–Karakoram mountain ranges in the northwestern region of the subcontinent. Consequently, this could account for the fact that reanalysis datasets overestimate more in regions with complex terrain than in flatter regions. Impressively, with regard to the seasonal and interannual variability of premonsoon precipitation over the mountainous regions, the newly released 3B42-V7 appears to provide useful information in comparison to CPC-uni, GPCP, and 3B42-V6.
For the winter season (DJF), all precipitation products display less agreement with respect to APHRODITE, illustrating important limitations in all the gridded precipitation, including our reference dataset. Based on the aforementioned reasons, a major source of uncertainty, particularly during the winter months over the data-sparse regions, is likely the representativeness and undercatch error in gauge measurements that tends to artificially inflate the existing errors in satellite and reanalyses precipitation estimates. In spite of the large winter biases in reanalyses (CFSR and ERA-Interim) and the newly released 3B42-V7, we are fairly confident that these precipitation products could be a useful source of information over the data-sparse regions. However, commenting on the accuracy and robustness of these products requires further investigation.
Nonetheless, the best approach to fill in these gaps is to recognize the relative strength and weaknesses of each precipitation product and to extract maximum possible information for their constructive applicability and usability. We anticipate that the outcomes of this study will benefit data users, model developers, and researchers of hydrology and climatology to understand the relevant seasonal characteristics of each precipitation product used in this study.
In closing, we would like to mention that future work will address the spatiotemporal variations of winter precipitation in the northwestern region of the Indian subcontinent using additional datasets and variables.
Acknowledgments
The authors would like to extend their sincere thanks to the Victoria University of Wellington (VUW), New Zealand, for supporting this research. The lead author acknowledges VUW for the Doctoral Scholarship Award, and the institutions/teams responsible for the creation and publication of the APHRODITE, CPC-uni, GPCP, TRMM, CFSR, and ERA-Interim data archives. We would also like to thank Dr. Joe Turk and the two anonymous reviewers for their constructive comments. The lead author also extends special thanks to Dr. Ankita Singh of the Indian Institute of Technology Bhubaneswar for her inputs regarding statistical techniques.
REFERENCES
Adler, R. F., and Coauthors, 2003: The Version-2 Global Precipitation Climatology Project (GPCP) Monthly Precipitation Analysis (1979–present). J. Hydrometeor., 4, 1147–1167, doi:10.1175/1525-7541(2003)004<1147:TVGPCP>2.0.CO;2.
Amitai, E., Llort X. , Liao L. , and Meneghini R. , 2004: A framework for global verification of space-borne radar estimates of precipitation based on rain type classification. 2nd TRMM Int. Science Conf., Nara, Japan, JAXA/NICT/NASA, 4 pp. [Available online at www.eorc.jaxa.jp/TRMM/museum/event/2ndTISC/HP/Extended%20Abstract/5.4_AMITAI_Eyal.pdf.]
Andermann, C., Bonnet S. , and Gloaguen R. , 2011: Evaluation of precipitation datasets along the Himalayan front. Geochem. Geophys. Geosyst., 12, Q07023, doi:10.1029/2011GC003513.
Barlow, M., Cullen H. , and Bradfield L. , 2002: Drought in central and southwest Asia: La Niña, the warm pool, and Indian Ocean precipitation. J. Climate, 15, 697–700, doi:10.1175/1520-0442(2002)015<0697:DICASA>2.0.CO;2.
Bookhagen, B., and Burbank D. W. , 2010: Toward a complete Himalayan hydrological budget: Spatiotemporal distribution of snowmelt and rainfall and their impact on river discharge. J. Geophys. Res., 115, F03019, doi:10.1029/2009JF001426.
Bosilovich, M. G., Chen J. , Robertson F. R. , and Adler R. F. , 2008: Evaluation of global precipitation in reanalyses. J. Appl. Meteor. Climatol, 47, 2279–2299, doi:10.1175/2008JAMC1921.1.
Bosilovich, M. G., Kennedy J. , Dee D. , Allan R. , and O’Neill A. , 2013: On the reprocessing and reanalysis of observations for climate. Climate Science for Serving Society: Research, Modeling and Prediction Priorities, G. R. Asrar and J. W. Hurrell, Eds., Springer, 51–71.
Brunetti, M., Colacino M. , Mugeri M. , and Nanni T. , 2001: Trends in the daily intensity of precipitation in Italy from 1951 to 1996. Int. J. Climatol., 21, 299–316, doi:10.1002/joc.613.
Chen, M., Shi W. , Xie P. , Silva V. B. S. , Kousky V. E. , Higgins R. W. , and Janowiak J. E. , 2008: Assessing objective techniques for gauge-based analyses of global daily precipitation. J. Geophys. Res., 113, D04110, doi:10.1029/2007JD009132.
Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597, doi:10.1002/qj.828.
Ding, Q. H., and Wang B. , 2005: Circumglobal teleconnection in the Northern Hemisphere summer. J. Climate, 18, 3483–3505, doi:10.1175/JCLI3473.1.
Dinku, T., Ceccato P. , Grover-Kopec E. , Lemma M. , Connor S. J. , and Ropelewski C. F. , 2007: Validation of satellite rainfall products over East Africa's complex topography. Int. J. Remote Sens., 28, 1503–1526, doi:10.1080/01431160600954688.
Ebert, E. E., Janowiak J. E. , and Kidd C. , 2007: Comparison of near-real-time precipitation estimates from satellite observations and numerical models. Bull. Amer. Meteor. Soc., 88, 47–64, doi:10.1175/BAMS-88-1-47.
Gebregiorgis, A. S., and Hossain F. , 2015: How well can we estimate error variance of satellite precipitation data across the world? Atmos. Res., 154, 39–59, doi:10.1016/j.atmosres.2014.11.005.
Goswami, B. N., and Ajaya Mohan R. S. , 2001: Intraseasonal oscillations and interannual variability of the Indian summer monsoon. J. Climate, 14, 1180–1198, doi:10.1175/1520-0442(2001)014<1180:IOAIVO>2.0.CO;2.
Goswami, B. N., Venugopal V. , Sengupta D. , Madhusoodanan M. S. , and Xavier P. K. , 2006: Increasing trend of extreme rain events over India in a warming environment. Science, 314, 1442–1445, doi:10.1126/science.1132027.
Gruber, A., Su X. , Kanamitsu M. , and Schemm J. , 2000: The comparison of two merged rain gauge–satellite precipitation datasets. Bull. Amer. Meteor. Soc., 81, 2631–2644, doi:10.1175/1520-0477(2000)081<2631:TCOTMR>2.3.CO;2.
Guhathakurta, P., and Rajeevan M. , 2008: Trends in the rainfall pattern over India. Int. J. Climatol., 28, 1453–1469, doi:10.1002/joc.1640.
Hamada, A., Arakawa O. , and Yatagai A. , 2011: An automated quality control method for daily rain-gauge data. Global Environ. Res.,15 (2), 183–192. [Available online at www.airies.or.jp/attach.php/6a6f75726e616c5f31352d32656e67/save/0/0/15_2-12.pdf.]
Horel, J. D., 1981: A rotated principal component analysis of the interannual variability of the Northern Hemisphere 500 mb height field. Mon. Wea. Rev., 109, 2080–2092, doi:10.1175/1520-0493(1981)109<2080:ARPCAO>2.0.CO;2.
Huffman, G. J., and Bolvin D. T. , 2014: TRMM and other data precipitation dataset documentation. NASA GSFC Rep., 42 pp. [Available online at ftp://precip.gsfc.nasa.gov/pub/trmmdocs/3B42_3B43_doc.pdf.]
Huffman, G. J., Morrissey M. , Bolvin D. T. , Curtis S. , Joyce R. , McGavock B. , and Susskind J. , 2001: Global precipitation at one-degree daily resolution from multisatellite observations. J. Hydrometeor., 2, 36–50, doi:10.1175/1525-7541(2001)002<0036:GPAODD>2.0.CO;2.
Huffman, G. J., and Coauthors, 2007: The TRMM Multi-satellite Precipitation Analysis: Quasi-global, multiyear, combined-sensor precipitation estimates at fine scale. J. Hydrometeor., 8, 38–55, doi:10.1175/JHM560.1.
Huffman, G. J., Bolvin D. T. , Nelkin E. J. , and Adler R. F. , 2011: Highlights of Version 7 TRMM Multi-satellite Precipitation Analysis (TMPA). Proc. 5th Int. Precipitation Working Group Workshop, Hamburg, Germany, Max-Planck-Institut für Meteorologie, 109–110. [Available online at www.mpimet.mpg.de/fileadmin/publikationen/Reports/WEB_BzE_100.pdf.]
Immerzeel, W. W., Droogers P. , de Jong S. M. , and Bierkens M. F. P. , 2009: Large-scale monitoring of snow cover and runoff simulation in Himalayan river basins using remote sensing. Remote Sens. Environ., 113, 40–49, doi:10.1016/j.rse.2008.08.010.
Joshi, M. K., Rai A. , and Pandey A. , 2012: Validation of TMPA and GPCP 1DD against the ground truth rain-gauge data for Indian region. Int. J. Climatol., 33, 2633–2648, doi:10.1002/joc.3612.
Kar, S. C., and Rana S. , 2013: Interannual variability of winter precipitation over northwest India and adjoining region: Impact of global forcings. Theor. Appl. Climatol., 116, 609–623, doi:10.1007/s00704-013-0968-z.
Kidd, C., Bauer P. , Turk J. , Huffman G. J. , Joyce R. , Hsu K.-L. , and Braithwaite D. , 2012: Intercomparison of high-resolution precipitation products over northwest Europe. J. Hydrometeor., 13, 67–83, doi:10.1175/JHM-D-11-042.1.
Kripalani, R. H., and Kumar P. , 2004: Northeast monsoon variability over south peninsular India vis-à-vis the Indian Ocean dipole mode. Int. J. Climatol., 24, 1267–1282, doi:10.1002/joc.1071.
Krishnamurthy, V., and Shukla J. , 2000: Intraseasonal and interannual variability of rainfall over India. J. Climate, 13, 4366–4377, doi:10.1175/1520-0442(2000)013<0001:IAIVOR>2.0.CO;2.
Kucera, P. A., and Coauthors, 2013: Precipitation from space: Advancing Earth system science. Bull. Amer. Meteor. Soc., 94, 365–375, doi:10.1175/BAMS-D-11-00171.1.
Kuligowski, R. J., 1997: An overview of National Weather Service quantitative precipitation estimates. TDL Office Note 97-4, NOAA, Silver Spring, MD, 27 pp. [Available online at www.nws.noaa.gov/im/pub/tdl97-4.pdf.]
Kumar, P., Rupa Kumar K. , Rajeevan M. , and Sahai A. K. , 2007: On the recent strengthening of the relationship between ENSO and northeast monsoon rainfall over South Asia. Climate Dyn,28, 649–660, doi:10.1007/s00382-006-0210-0.
Lorenz, C., and Kunstmann H. , 2012: The hydrological cycle in three state-of-the-art reanalyses: Intercomparison and performance analysis. J. Hydrometeor., 13, 1397–1420, doi:10.1175/JHM-D-11-088.1.
Ma, L., Zhang T. , Frauenfeld W. O. , Ye B. , Yang D. , and Qin D. , 2009: Evaluation of precipitation from the ERA-40, NCEP-1, and NCEP-2 reanalyses and CMAP-1, CMAP-2, and GPCP-2 with ground-based measurements in China. J. Geophys. Res., 114, D09105, doi:10.1029/2008JD011178.
Mahanta, R., Sama D. , and Choudhury A. , 2013: Heavy rainfall occurrences over northeast India. Int. J. Climatol., 33, 1456–1469, doi:10.1002/joc.3526.
Mishra, V., Smoliak B. V. , Lettenmaier D. P. , and Wallace J. M. , 2012: A prominent pattern of year-to-year variability in Indian summer monsoon rainfall. Proc. Natl. Acad. Sci. USA, 109, 7213–7217, doi:10.1073/pnas.1119150109.
Misra, V., Pantina P. , Chan S. C. , and DiNapoli S. , 2012: A comparative study of the Indian monsoon hydroclimate and its variations in three reanalyses. Climate Dyn., 39, 1149–1168, doi:10.1007/s00382-012-1319-y.
Nair, A., Acharya N. , Singh A. , Mohanty U. C. , and Panda T. C. , 2013: On the predictability of northeast monsoon rainfall over south peninsular India in general circulation models. Pure Appl. Geophys., 170, 1945–1967, doi:10.1007/s00024-012-0633-y.
New, M., Todd M. , Hulme M. , and Jones P. , 2001: Precipitation measurements and trends in the twentieth century. Int. J. Climatol., 21, 1889–1922, doi:10.1002/joc.680.
Palazzi, E., Hardenberg J. V. , and Provenzale A. , 2013: Precipitation in the Hindu-Kush Karakoram Himalaya: Observations and future scenarios. J. Geophys. Res. Atmos., 118, 85–100, doi:10.1029/2012JD018697.
Pattanaik, D. R., and Rajeevan M. , 2010: Variability of extreme rainfall events over India during southwest monsoon season. Meteor. Appl., 17, 88–104, doi:10.1002/met.164.
Petty, G. W., and Krajewski W. , 1996: Satellite rainfall estimation over land. Hydrol. Sci. J., 41, 433–451, doi:10.1080/02626669609491519.
Prakash, S., and Coauthors, 2015: Seasonal intercomparison of observational rainfall datasets over India during the southwest monsoon season. Int. J. Climatol., doi:10.1002/joc.4129, in press.
Rahman, S. H., Sengupta D. , and Ravichandran M. , 2009: Variability of Indian summer monsoon rainfall in daily data from gauge and satellite. J. Geophys. Res., 114, D17113, doi:10.1029/2008JD011694.
Rajeevan, M., and Bhate J. , 2009: A high resolution daily gridded rainfall dataset (1971–2005) for mesoscale meteorological studies. Curr. Sci., 96, 558–562.
Rajeevan, M., Gadgil S. , and Bhate J. , 2010: Active and break spells of the Indian summer monsoon. J. Earth Syst. Sci., 119, 229–247, doi:10.1007/s12040-010-0019-4.
Rajeevan, M., Unnikrishnan C. K. , Bhate J. , Kumar N. K. , and Sreekala P. P. , 2012: Northeast monsoon over India: Variability and prediction. Meteor. Appl., 19, 226–236, doi:10.1002/met.1322.
Rasmussen, R., and Coauthors, 2012: How well are we measuring snow? The NOAA/FAA/NCAR winter precipitation test bed. Bull. Amer. Meteor. Soc., 93,811–829, doi:10.1175/BAMS-D-11-00052.1.
Richards, F., and Arkin P. A. , 1981: On the relationship between satellite-observed cloud cover and precipitation. Mon. Wea. Rev., 109, 1081–1093, doi:10.1175/1520-0493(1981)109<1081:OTRBSO>2.0.CO;2.
Richman, M. B., 1986: Rotation of principal components. J. Climatol., 6, 293–335, doi:10.1002/joc.3370060305.
Rudolf, B., and Rubel F. , 2005: Global precipitation. Observed Global Climate, M. Hantel, Ed., Landolt-Börnstein: Numerical Data and Functional Relationships in Science and Technology, Vol. 6, Springer-Verlag, 11.01–11.53.
Saha, S., and Coauthors, 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 1015–1057, doi:10.1175/2010BAMS3001.1.
Sen Roy, S. S., 2006: The impacts of ENSO, PDO, and local SSTs on winter precipitation in India. Phys. Geogr., 27, 464–474, doi:10.2747/0272-3646.27.5.464.
Sen Roy, S. S., 2009: A spatial analysis of extreme hourly precipitation patterns in India. Int. J. Climatol., 29, 345–355, doi:10.1002/joc.1763.
Sevruk, B., 1985: Correction of precipitation measurements: Summary report. Proc. Workshop on the Correction of Precipitation Measurements, Zurich, Switzerland, ETH/IAHS/WMO, 13–23.
Shah, R., and Mishra V. , 2014: Evaluation of the reanalysis products for the monsoon season droughts in India. J. Hydrometeor., 15, 1575–1591, doi:10.1175/JHM-D-13-0103.1.
Shea, D. J., and Sontakke N. A. , 1995: The annual cycle of precipitation over the Indian subcontinent: Daily, monthly and seasonal statistics. NCAR Tech. Note NCAR/TN-401+STR, 168 pp., doi:10.5065/D65M63M8.
Shen, Y., Xiong A. , Wang Y. , and Xie P. , 2010: Performance of high-resolution satellite precipitation products over China. J. Geophys. Res., 115, D02114, doi:10.1029/2009JD012097.
Sohn, S. J., Tam C. Y. , Ashok K. , and Ahn J. B. , 2012: Quantifying the reliability of precipitation datasets for monitoring large-scale East Asian precipitation variations. Int. J. Climatol., 32, 1520–1526, doi:10.1002/joc.2380.
Srinivasan, V., and Ramamurthy K. , 1973: Northeast monsoon. Forecasting manual part IV: Comprehensive articles on selected topics, FMU Rep. IV-18.4, IMD, Pune, India. [Available online at www.imdpune.gov.in/weather_forecasting/Forecasting_Mannuals/IMD_IV-18.4.pdf.]
Syed, F. S., Giorgi F. , Pal J. S. , and Keay K. , 2009: Regional climate model simulation of winter climate over central–Southwest Asia, with emphasis on NAO and ENSO effects. Int. J. Climatol., 30, 220–235, doi:10.1002/joc.1887.
Trenberth, K. E., 2011: Changes in precipitation with climate change. Climate Res., 47, 123–138, doi:10.3354/cr00953.
Trenberth, K. E., Fasullo J. T. , and Mackaro J. , 2011: Atmospheric moisture transports from ocean to land and global energy flows in reanalyses. J. Climate, 24, 4907–4924, doi:10.1175/2011JCLI4171.1.
TRMM Precipitation Radar Team, 2011: Tropical Rainfall Measuring Mission (TRMM) Precipitation Radar algorithm instruction manual for version 7. Tech. Rep., JAXA/NASA, 170 pp. [Available online at www.eorc.jaxa.jp/TRMM/documents/PR_algorithm_product_information/pr_manual/PR_Instruction_Manual_V7_L1.pdf.]
Turk, F. J., Arkin P. A. , Ebert E. E. , and Sapiano M. R. P. , 2008: Evaluating high-resolution precipitation products. Bull. Amer. Meteor. Soc., 89, 1911–1916, doi:10.1175/2008BAMS2652.1.
Turner, A. G., and Annamalai H. , 2012: Climate change and the South Asian summer monsoon. Nat. Climate Change, 2, 587–595, doi:10.1038/nclimate1495.
Wang, B., and Ding Q. H. , 2006: Changes in global monsoon precipitation over the past 56 years. Geophys. Res. Lett., 33, L06711, doi:10.1029/2005GL025347.
Wilks, D. S., 1995: Statistical Methods in Atmospheric Sciences. International Geophysics Series, Vol. 59, Academic Press, 464 pp.
Xie, P., and Coauthors, 2003: GPCP pentad precipitation analyses: An experimental dataset based on gauge observations and satellite estimates. J. Climate, 16, 2197–2214, doi:10.1175/2769.1.
Yao, T., and Coauthors, 2012: Different glacier status with atmospheric circulations in Tibetan Plateau and surroundings. Nat. Climate Change, 2, 663–667, doi:10.1038/nclimate1580.
Yatagai, A., and Xie P. , 2006: Utilization of a rain-gauge-based daily precipitation dataset over Asia for validation of precipitation derived from TRMM/PR and JRA25. Remote Sensing and Modeling of the Atmosphere, Oceans, and Interactions, T. N. Krishnamurti, B. N. Goswami, and T. Iwasaki, Eds., International Society for Optical Engineering (SPIE Proceedings, Vol. 0604), 64040M, doi:10.1117/12.723829.
Yatagai, A., Xie P. , and Alpert P. , 2008: Development of a daily gridded precipitation dataset for the Middle East. Adv. Geosci., 12, 165–170, doi:10.5194/adgeo-12-165-2008.
Yatagai, A., Arakawa O. , Kamiguchi K. , Kawamoto H. , Nodzu M. I. , and Hamada A. , 2009: A 44-year daily gridded precipitation dataset for Asia based on a dense network of rain gauges. SOLA, 5, 137–140, doi:10.2151/sola.2009-035.
Yatagai, A., Kamiguchi K. , Arakawa O. , Hamada A. , Yasutomi N. , and Kitoh A. , 2012: APHRODITE: Constructing a long-term daily gridded precipitation dataset for Asia based on a dense network of rain gauges. Bull. Amer. Meteor. Soc., 93, 1401–1415, doi:10.1175/BAMS-D-11-00122.1.
Yilmaz, K., Hogue T. S. , Hsu K.-L. , Sorooshian S. , Gupta H. V. , and Wagener T. , 2005: Evaluation of rain gauge, radar and satellite-based precipitation estimates with emphasis on hydrologic forecasting. J. Hydrometeor., 6, 497–517, doi:10.1175/JHM431.1.
Zhang, Q., Kornich H. , and Holmgren K. , 2013: How well do reanalyses represent the southern African precipitation? Climate Dyn., 40, 951–962, doi:10.1007/s00382-012-1423-z.
