Relative impact of observations on a regional Arctic numerical weather prediction system

Abstract

The impact of Arctic conventional and satellite observations on regional short-range weather forecasts is assessed using observing-system experiments, in which observations are removed (denied) when creating the initial conditions of the forecasts. The experiments are conducted with the AROME-Arctic regional mesoscale numerical weather prediction system, using as lateral boundary conditions (LBCs) observing-system experiments performed at the European Centre for Medium-Range Forecasts (ECMWF) with the global forecasting system. This allows the assessment of the relative impacts of observations on forecast skill through regional data assimilation (DA), through LBCs, and the total impact due to the denial of observations in both the regional and global forecasting systems. The study is conducted during the first and second Special Observing Periods of the Year Of Polar Prediction. The total impact on the upper-air forecasts is dominated by the impact of observations through their assimilation in the LBCs, while for the winter period the impact on surface fields is dominated by the regional DA. The latter is significant up to 36 hr, while the former impact can last throughout the verified forecast range (48 hr). The use of observations in the LBCs has both significantly positive and significantly negative impacts. In terms of total impact on forecast skill, conventional observations, followed by infrared radiances, have the largest impact on all upper-air parameters, except for humidity. For upper-air humidity forecasts, the microwave radiances have the largest impact. In terms of observation impact through regional DA, conventional observations play the largest role for upper-air temperature and geopotential, microwave for upper-air humidity, and atmospheric motion vectors and Infrared Atmospheric Sounding Interferometer (IASI) for wind forecast. Regional DA of conventional observations also contributes the most to improvements of surface fields, except for 10-m winds, for which the microwave temperature-sensitive radiances are the most important.

On the configuration of a regional Arctic Numerical Weather Prediction system to maximize predictive capacity

“…For RH2m a modest reduction (1-7.5%) in forecast skill is found for day 1 for inland, coast and fjords. The results presented here are in line with aggregated results for the entire domain from Randriamampianina et al (2021), who also explain the deterioration in RH2m quality by an unbalanced adjustment of temperature and specific humidity in the assimilation process.…”

“…It is also well documented that weather forecasts produced by the regional Arctic NWP system are sensitive to the initialization of the model atmosphere and surface (e.g. Køltzow et al, 2019;Randriamampianina et al, 2019Randriamampianina et al, , 2021. The forecast quality is substantially improved by assimilation of observations in this specific region compared to simply starting regional integrations from coarser spatial resolution global model analyses.…”

Evaluating the use of Aeolus satellite observations in the regional numerical weather prediction (NWP) model Harmonie–Arome

“…Since the launch on 22 August 2018 it has been orbiting the Earth in a sun-synchronous orbit at 320 km of height, providing vertical wind speed profiles measured with a Doppler wind lidar. A Doppler wind lidar is an active instrument, and it derives wind measurements by detecting the shift in the backscatter signal from the onboard laser using an instrument called ALADIN (Atmospheric Laser Doppler Instrument; see Reitebuch et al, 2009). The winds derived from the satellite S. Hagelin et al: Evaluating the use of Aeolus satellite observations measurements are perpendicular to the direction of travel, hereafter referred to as the HLOS (horizontal line-of-sight) winds.…”

The potential of numerical prediction systems to support the design of Arctic observing systems: Insights from the APPLICATE and YOPP projects

“…OSEs were performed for several seasons, among which the YOPP Special Observing Periods for the Northern Hemisphere (February to March and June to August 2018), and the analysis focused mainly on the impact of Arctic observations. The results highlighted the added value of Arctic observations for short‐ and medium‐range forecast skill (Lawrence et al ., 2019a; Laroche and Poan, 2021; Randriamampianina et al ., 2021) by showing that: In global NPS, all current Arctic atmospheric observing systems increase short‐ and medium‐range predictive skill both in polar regions and midlatitudes; In all the contributing global NPS, conventional Arctic observations have the largest impact in winter, while in summer the leading observing system varies from one forecasting system to another: radiances from microwave sounders play the biggest role in the ECMWF system, while conventional observations are most important at ECCC and DWD. This demonstrates that observation impacts are always subject to the sophistication/maturity of the data assimilation system, the forecast model and of the assessment and monitoring of the observations' quality; The use of microwave sounder observations in the ECMWF IFS is suboptimal during winter.…”