Covariance inflation plays an important role in the ensemble Kalman filter because the ensemble‐based error variance is usually underestimated due to various factors such as the limited ensemble size and model imperfections. Manual tuning of the inflation parameters by trial and error is computationally expensive; therefore, several studies have proposed approaches to adaptive estimation of the inflation parameters. Among others, this study focuses on the covariance relaxation method which realizes spatially dependent inflation with a spatially homogeneous relaxation parameter. This study performs a series of experiments with the non‐hydrostatic icosahedral atmospheric model (NICAM) and the local ensemble transform Kalman filter (LETKF) assimilating the real‐world conventional observations and satellite radiances. Two adaptive covariance relaxation methods are implemented: relaxation to prior spread based on Ying and Zhang (adaptive‐RTPS), and relaxation to prior perturbation (adaptive‐RTPP). Both adaptive‐RTPS and adaptive‐RTPP generally improve the analysis compared to a baseline control experiment with an adaptive multiplicative inflation method. However, the adaptive‐RTPS and adaptive‐RTPP methods lead to an over‐dispersive (under‐dispersive) ensemble in the sparsely (densely) observed regions compared with the adaptive multiplicative inflation method. We find that the adaptive‐RTPS and adaptive‐RTPP methods are robust to a sudden change in the observing networks and observation error settings.

Japan's new geostationary satellite “Himawari‐8” enables updating precipitation and flood predictions as frequently as every 10 min and to capture as early as possible the flood risk associated with rapidly changing severe weather. This study focuses on the advantage of the frequent update of precipitation and flood predictions by assimilating all‐sky Himawari‐8 infrared radiances in the case of September 2015 Kanto‐Tohoku heavy rainfall, a major flooding event in Japan. The analyzed tropical cyclone representation and moisture transport are improved by the Himawari‐8 data assimilation. Deterministic runs from the analyses with the Himawari‐8 data provide much improved forecasts of a strong precipitation band. Moreover, every 10 min updates of river discharge forecasts driven by these improved precipitation forecasts show a large improvement in forecast skill with longer lead times. The results demonstrate that the frequent refresh aids to capture a signature of flooding at an earlier stage.

This study aims to propose two new approaches to improve precipitation forecasts from numerical weather prediction (NWP) models through effective data assimilation of satellite‐derived precipitation. The assimilation of precipitation data is known to be very difficult mainly because of highly non‐Gaussian statistics of precipitation variables. Following Lien et al., this study addresses the non‐Gaussianity issue by applying the Gaussian transformation (GT) based on the empirical cumulative distribution function (CDF) of precipitation. We propose a method that constructs the CDF with only recent 1 month samples, without using a long period of samples needed previously. We also propose a method to use the inverse GT, with which we can obtain realistic precipitation fields from biased NWP model outputs. We assimilate the Japan Aerospace eXploration Agency's Global Satellite Mapping of Precipitation (GSMaP) data into the Nonhydrostatic Icosahedral Atmospheric Model (NICAM) at 112 km horizontal resolution. Assimilating the GSMaP data results in improved weather forecasts compared to the control experiment assimilating only rawinsonde data. We find that horizontal observation thinning is necessary, probably due to the horizontal observation‐error correlations in the GSMaP data. We also obtained precipitation fields similar to GSMaP from the NICAM precipitation forecasts by using the inverse GT, leading to an improved precipitation forecast.

To date, many studies have performed numerical estimations of biomass production and agricultural water demand to understand the present and future supply–demand relationship. A crop calendar (CC), which defines the date or month when farmers sow and harvest crops, is an essential input for the numerical estimations. This study aims to present a new global data set, the SAtellite-derived CRop calendar for Agricultural simulations (SACRA), and to discuss advantages and disadvantages compared to existing census-based and model-derived products. We estimate global CC at a spatial resolution of 5 arcmin using satellite-sensed normalized difference vegetation index (NDVI) data, which corresponds to vegetation vitality and senescence on the land surface. Using the time series of the NDVI averaged from three consecutive years (2004–2006), sowing/harvesting dates are estimated for six crops (temperate-wheat, snow-wheat, maize, rice, soybean and cotton). We assume time series of the NDVI represent the phenology of one dominant crop and estimate CCs of the dominant crop in each grid. The dominant crops are determined using harvested areas based on census-based data. The cultivation period of SACRA is identified from the time series of the NDVI; therefore, SACRA considers current effects of human decisions and natural disasters. The difference between the estimated sowing dates and other existing products is less than 2 months (< 62 days) in most of the areas. A major disadvantage of our method is that the mixture of several crops in a grid is not considered in SACRA. The assumption of one dominant crop in each grid is a major source of discrepancy in crop calendars between SACRA and other products. The disadvantages of our approach may be reduced with future improvements based on finer satellite sensors and crop-type classification studies to consider several dominant crops in each grid. The comparison of the CC also demonstrates that identification of wheat type (sowing in spring or fall) is a major source of error in global CC estimations.