Uncertainties in calculating precipitation climatology in East Asia

Abstract

Abstract. This study examines the uncertainty in calculating the fundamental climatological characteristics of precipitation in the East Asia region from multiple fine-resolution gridded analysis data sets based on in situ rain gauge observations and data assimilations. Five observation-based gridded precipitation data sets are used to derive the long-term means, standard deviations in lieu of interannual variability and linear trends over the 28-year period from 1980 to 2007. Both the annual and summer (June-July-August) mean precipitation is examined. The agreement amongst these precipitation data sets is examined using two metrics including the signal-to-noise ratio (SNR) defined as the ratio between long-term means and the corresponding standard deviations, and Taylor diagrams, which allow examinations of the pattern correlation, the standard deviation, and the centered root mean square error. It is found that the five gauge-based precipitation analysis data sets agree well in the long-term mean and interannual variability in most of the East Asia region including eastern China, Manchuria, South Korea, and Japan, which are densely populated and have fairly highdensity observation networks. The regions of large interdata-set variations include Tibetan Plateau, Mongolia, northern Indo-China, and North Korea. The regions of large uncertainties are typically lightly populated and are characterized by severe terrain and/or extremely high elevations. Unlike the long-term mean and interannual variability, agreement between data sets in the linear trend is weak, both for the annual and summer mean values. In most of the East Asia region, the SNR for the linear trend is below 0.5: the interdata-set variability exceeds the multi-data ensemble mean.The uncertainty in the spatial distribution of long-term means among these data sets occurs both in the spatial pattern and variability, but the uncertainty for the interannual variability and time trend is much larger in the variability than in the pattern correlation. Thus, care must be taken in using long-term trends calculated from gridded precipitation analysis data for climate studies over the East Asia region.

Intercomparison of precipitation datasets for summer precipitation characteristics over East Asia

“…This may be due to the reanalysis models' inability to represent the effects of complex orography and/or sparse observational inputs for assimilations. It must be noted that the observed CMAP and GPCP data also suffer from the lack of gauge data in these regions (e.g., Kim et al 2015a;Kim and Park 2016). For higher modes (Fig.…”

“…Previous studies (e.g., Kim et al 2015a;Kim and Park 2016) showed that the differences between observation-based gridded precipitation datasets vary widely according to regions, seasons, precipitation characteristics and/or statistical properties. Overall, the long-term means and the temporal standard deviation over the 37 summers, a surrogate for the seasonal cycle and its interannual variability respectively, of the GPCP data agree closely with the CMAP data ( Table 2).…”

“…Kim et al (2015a) revealed that the spread among observation-based gridded precipitation analyses varies seasonally as well as regionally. Kim and Park (2016) also showed that the spread among gridded precipitation datasets over East Asia varies according to statistical properties, i.e., smallest in seasonal means and largest in longterm trends, in addition to geography and season. Such differences arise from various sources involved in analysis processes such as input data quality control, observation density, analysis schemes, and, for reanalysis datasets, the differences in model formulations (Legates and Willmott 1990;Kim et al 2015a;Kim and Park 2016).…”

“…Previous studies (Prakash et al 2014;Kim et al 2015a;Kim and Park 2016) found that precipitation characteristics in observation-based gridded datasets often vary widely among them. This is an important concern because model evaluation against observational and/or reanalysis data is critical for model developments/improvements and in the application of climate model data to impact assessments via bias correction (Kim et al 2015a), for example.…”

Ensemble Spatial Precipitation Analysis From Rain Gauge Data: Methodology and Application in the European Alps

“…Also, its stochastic foundation provides some confidence that the uncertainty information is statistically reliable. Our experiments (section 5) pinpoint to the risk of underestimating uncertainty considerably with the popular ad hoc comparison of a set of deterministic results (e.g., Kim & Park, 2016;Kotlarski et al, 2017;Prein & Gobiet, 2016). The members of such a pseudo-ensemble all represent a sort of optimal estimate, based on mostly the same observations, rather than independent realizations (see also Chappell et al, 2012).…”