NASA Goddard Earth Observing System Research Articles

Contributions of Initial Conditions and Meteorological Forecast to Subseasonal-to-Seasonal Hydrological Forecast Skill in Western Tropical South America

Abstract Hydrological predictions at subseasonal-to-seasonal (S2S) time scales can support improved decision-making in climate-dependent sectors like agriculture and hydropower. Here, we present an S2S hydrological forecasting system (S2S-HFS) for western tropical South America (WTSA). The system uses the global NASA Goddard Earth Observing System S2S meteorological forecast system (GEOS-S2S) in combination with the generalized analog regression downscaling algorithm and the NASA Land Information System (LIS). In this implementation study, we evaluate system performance for 3-month hydrological forecasts for the austral autumn season (March–May) using ensemble hindcasts for 2002–17. Results indicate that the S2S-HFS generally offers skill in predictions of monthly precipitation up to 1-month lead, evapotranspiration up to 2 months lead, and soil moisture content up to 3 months lead. Ecoregions with better hindcast performance are located either in the coastal lowlands or in the Amazon lowland forest. We perform dedicated analysis to understand how two important teleconnections affecting the region are represented in the S2S-HFS: El Niño–Southern Oscillation (ENSO) and the Antarctic Oscillation (AAO). We find that forecast skill for all variables at 1-month lead is enhanced during the positive phase of ENSO and the negative phase of AAO. Overall, this study indicates that there is meaningful skill in the S2S-HFS for many ecoregions in WTSA, particularly for long memory variables such as soil moisture. The skill of the precipitation forecast, however, decays rapidly after forecast initialization, a phenomenon that is consistent with S2S meteorological forecasts over much of the world.

Improved simulations of biomass burning aerosol optical properties and lifetimes in the NASA GEOS Model during the ORACLES-I campaign

Abstract. In order to improve aerosol representation in the NASA Goddard Earth Observing System (GEOS) model, we evaluated simulations of the transport and properties of aerosols from southern African biomass burning sources that were observed during the first deployment of the NASA ORACLES (ObseRvations of Aerosols above CLouds and their intEractionS) field campaign in September 2016. An example case study of 24 September was analyzed in detail, during which aircraft-based in situ and remote sensing observations showed the presence of a multi-layered smoke plume structure with significant vertical variation in single scattering albedo (SSA). Our baseline GEOS simulations were not able to represent the observed SSA variation or the observed organic aerosol-to-black-carbon ratio (OA : BC). Analyzing the simulated smoke age suggests that the higher-altitude, less absorbing smoke plume was younger (∼4 d), while the lower-altitude and more absorbing smoke plume was older (∼7 d). We hypothesize a chemical or microphysical loss process exists to explain the change in aerosol absorption as the smoke plume ages, and we apply a simple loss rate to the model hydrophilic biomass burning OA to simulate this process. We also utilized the ORACLES airborne observations to better constrain the simulation of aerosol optical properties, adjusting the assumed particle size, hygroscopic growth, and absorption. Our final GEOS model simulation with additional OA loss and updated optics showed better performance in simulating aerosol optical depth (AOD) and SSA compared to independent ground- and space-based retrievals for the entire month of September 2016, including the Ozone Monitoring Instrument (OMI) Aerosol Index. In terms of radiative implications of our model adjustments, the final GEOS simulation suggested a decreased atmospheric warming of about 10 % (∼2 W m−2) over the southeastern Atlantic region and above the stratocumulus cloud decks compared to the model baseline simulations. These results improve the representation of the smoke age, transport, and optical properties in Earth system models.

Investigation of observed dust trends over the Middle East region in NASA Goddard Earth Observing System (GEOS) model simulations

Abstract. Satellite observations and ground-based measurements have indicated a high variability in the aerosol optical depth (AOD) in the Middle East region in recent decades. In the period that extends from 2003 to 2012, observations show a positive AOD trend of 0.01–0.04 per year or a total increase of 0.1–0.4 per decade. This study aimed to investigate if the observed trend was also captured by the NASA Goddard Earth Observing System (GEOS) model. To this end, we examined changes in the simulated dust emissions and dust AOD during this period. To understand the factors driving the increase in AOD in this region we also examined meteorological and surface parameters important for dust emissions, such as wind fields and soil moisture. Two GEOS model simulations were used in this study: the Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2) reanalysis (with meteorological and aerosol AOD data assimilated) and MERRA-2 Global Modeling Initiative (GMI) Replay (with meteorology constrained by the MERRA-2 reanalysis but without aerosol assimilation). We did not find notable changes in the modeled 10 m wind speed and soil moisture. However, analysis of Moderate Resolution Imaging Spectroradiometer (MODIS) normalized difference vegetation index (NDVI) data did show an average decrease of 8 % per year in the region encompassing Syria and Iraq, which prompted us to quantify the effects of vegetation on dust emissions and AOD in the Middle East region. This was done by performing a sensitivity experiment in which we enhanced dust emissions in grid cells where the NDVI decreased. The simulation results supported our hypothesis that the loss of vegetation cover and the associated increase in dust emissions over Syria and Iraq can partially explain the increase in AOD downwind. The model simulations indicated dust emissions need to be 10-fold larger in those grid cells in order to reproduce the observed AOD and trend in the model.

Adapting subseasonal-to-seasonal (S2S) precipitation forecast at watersheds for hydrologic ensemble streamflow forecasting with a machine learning-based post-processing approach

Accurate and reliable precipitation predictions made by dynamical forecast models could provide crucial information for human socioeconomic activities by enabling hydrologic forecasts at the Subseasonal-to-Seasonal (S2S) timescale. To utilize available S2S precipitation predictions for hydrologic forecasts, post-processing techniques have been applied to adapt the raw S2S precipitation to local watersheds. However, conventional statistical-based post-processing techniques are more focused on correcting the forecast bias, but rather limited in improving the predictive skill of available S2S precipitation forecasts. In this study, we combine the Random Forest classifiers (RF) with the Bias Correction and Spatial Disaggregation (BCSD) to adapt the 10-member ensemble precipitation forecast from the NASA Goddard Earth Observing System model version 5 (GEOS5) at 4 watersheds located in the NCEI South climate region of the United States. The adapted S2S precipitation is further applied for streamflow forecast by forcing a classical lumped hydrologic model. The performance of S2S precipitation as well as the corresponding streamflow predictions are benchmarked with the randomly resampled precipitation and the corresponding Ensemble Streamflow Prediction (ESP) framework-generated streamflow predictions. Evaluation statistics of Kling-Gupta Efficiency (KGE), Continuous Ranked Probability Skill Score (CRPSS), Reliability, Resolution, and Sharpness are employed to evaluate the predictive skill of precipitation and streamflow both deterministically and probabilistically. Our results indicate that dynamical S2S precipitation after forecast adaptation leads to consistently higher deterministic skill over ESP at all forecast lead times and across study watersheds. However, at longer forecast lead times beyond 10–15 days, S2S precipitation with a limited ensemble size does not present higher probabilistic skill than ESP. Our results shows that the joint application of RF and BCSD improves the predictive skill of the raw S2S precipitation at study watersheds in contrast to BCSD. Further, the added predictive skill of S2S precipitation brought by RF propagates into streamflow predictions, predominantly at longer forecast lead times exceeding 10 days. Overall, our results highlight the potential success of future work to apply other data-driven approaches to adapt the raw precipitation to local watersheds for more accurate and reliable streamflow forecasts at the S2S timescale.

Direct assimilation of AVHRR satellite radiance measurements in a reanalysis system

AbstractThe Advanced Very‐High‐Resolution Radiometer (AVHRR) is a broad‐band, five‐channel scanner sensing in the visible, near‐infrared and thermal infrared portions of the electromagnetic spectrum. AVHRR instruments onboard polar‐orbiting satellites have data records spanning 30 years. The radiances from the infrared channels of AVHRR have been directly assimilated over the ice‐free ocean in the NASA Goddard Earth Observing System (GEOS) since 2017 to constrain skin sea surface temperature (SST). The GEOS system already uses an advanced bulk SST and a skin SST model which incorporates diurnal warming and cool‐skin temperature on the surface of the ocean. The positive contribution of this effort to the Numerical Weather Prediction (NWP) makes it desirable to extend the skin SST data assimilation procedure to reanalysis applications. The AVHRR data from platforms that were previously unexplored for NWP applications are assimilated at the Global Modeling and Assimilation Office (GMAO) at NASA, for an upcoming reanalysis project. This study uses results from reanalysis experiments to assess the impact of assimilating AVHRR radiances over open ocean on the atmospheric state variables, focusing on the analyzed skin temperatures that include the active skin SST model. It is demonstrated that including the skin temperature variations in atmospheric analysis is generally beneficial. The addition of radiance measurements from AVHRR provides further improvements to the overall reanalysis system performance. by helping to correct the diurnal heating and reducing errors in the background departure of hyperspectral radiance observations, which are an essential component of atmospheric reanalyses.

Tropospheric ozone data assimilation in the NASA GEOS Composition Forecast modeling system (GEOS-CF v2.0) using satellite data for ozone vertical profiles (MLS), total ozone columns (OMI), and thermal infrared radiances (AIRS, IASI)

The NASA Goddard Earth Observing System Composition Forecast system (GEOS-CF) provides global near-real-time analyses and forecasts of atmospheric composition. The current version of GEOS-CF builds on the GEOS general circulation model with Forward Processing assimilation of meteorological data (GEOS-FP) and includes detailed GEOS-Chem tropospheric and stratospheric chemistry. Here we add 3D variational data assimilation in GEOS-CF to assimilate satellite observations of ozone including MLS vertical profiles, OMI total columns, and AIRS and IASI hyperspectral 9.6 μm radiances. We focus our evaluations on the troposphere. We find that the detailed tropospheric chemistry in GEOS-CF significantly improves the simulated background ozone fields relative to previous versions of the GEOS model, allowing for specification of smaller background errors in assimilation and resulting in smaller assimilation increments to correct the simulated ozone. Assimilation increments are largest in the upper troposphere and are consistent between satellite data sets. The OMI and MLS ozone data generally provide more information than the AIRS and IASI radiances except at high latitudes where the radiances provide more information. Comparisons to independent ozonesonde and aircraft (ATom-4) observations for 2018 show significant GEOS-CF improvement from the assimilation, particularly in the extratropical upper troposphere.

Interactive Stratospheric Aerosol Microphysics‐Chemistry Simulations of the 1991 Pinatubo Volcanic Aerosols With Newly Coupled Sectional Aerosol and Stratosphere‐Troposphere Chemistry Modules in the NASA GEOS Chemistry‐Climate Model (CCM)

AbstractWe have coupled the GEOS‐Chem tropospheric‐stratospheric chemistry mechanism and the Community Aerosol and Radiation Model for Atmospheres (CARMA), a sectional aerosol microphysics module, within the NASA Goddard Earth Observing System Chemistry‐Climate Model (GEOS CCM) in order to simulate the interactions between stratospheric chemistry and aerosol microphysics. We use observations of the 1991 Mount Pinatubo volcanic cloud to evaluate this new version of the GEOS CCM. The GEOS‐Chem chemistry module is used to simulate the oxidation of sulfur dioxide (SO2) more realistically than assuming hydroxyl radical (OH) fields are constant, as OH concentrations in the plume decrease dramatically in the weeks following the eruption. CARMA simulates sulfate aerosols with dynamic microphysical and optical properties. The CARMA‐calculated aerosol surface area is coupled to the chemistry module from GEOS‐Chem for the calculation of heterogeneous chemistry. We use a set of observational and theoretical constraints for Pinatubo to evaluate the performance of this new version of the GEOS CCM. These simulations are specifically compared with satellite and in‐situ observations and provide insights into the connections between the gas‐phase chemistry and the aerosol microphysics of the early plume and how they impact the climatic and chemical changes following a large volcanic eruption. A second, smaller eruption is also included in these simulations, the 15 August 1991, eruption of Cerro Hudson in Chile, which we find essential in explaining the aerosol optical depth in the Southern Hemisphere in 1991.

A weakly coupled land surface analysis with SMAP radiance assimilation improves GEOS medium‐range forecasts of near‐surface air temperature and humidity

AbstractThe NASA Goddard Earth Observing System (GEOS) employs a hybrid four‐dimensional ensemble‐variational (Hybrid‐4DEnVar) atmospheric data assimilation system to provide global, near‐real time weather analysis and forecast products. This study introduces a land analysis into the GEOS Hybrid‐4DEnVar system to additionally assimilate L‐band (1.4 GHz) brightness temperature observations over land from the NASA Soil Moisture Active Passive (SMAP) satellite mission, which are highly sensitive to surface (∼0–5 cm) soil moisture. This weakly coupled land analysis impacts the simulated and forecast atmosphere through the land–atmosphere exchange processes encoded in the atmospheric model. Retrospective assimilation experiments for boreal summer 2017 were conducted with the system at 50 km horizontal resolution. The SMAP assimilation is shown to mitigate errors in screen‐level (2 m) specific humidity (q2m) and temperature (T2m), with regional reductions in the root‐mean‐square error (RMSE) of q2m and maximum daily T2m by up to 0.4 g·kg−1 and 0.3 K, respectively. These improvements are somewhat smaller than those found in a precursor study that used the atmospheric analysis in the simpler, 3‐dimensional variational (3DVar) atmospheric assimilation configuration of the current GEOS reanalysis; Hybrid‐4DEnVar provides overall better near‐surface background estimates than 3DVar and therefore leaves less room for improvement. Finally, in the Hybrid‐4DEnVar system with SMAP assimilation, forecasts of q2m and T2m have significantly improved anomaly correlation and RMSE (99% confidence) at lead times out to 5 days compared to the Hybrid‐4DEnVar system without SMAP assimilation, with medium‐range lead times extended by ∼3 hr for q2m and ∼2 hr for T2m. Slight but nevertheless significant improvements are also seen for temperature forecasts at the 925 and 850 hPa levels and lead times out to 4 days. Humidity forecasts at 925 and 850 hPa are improved out to 1.5‐day lead time with 90% confidence.

Simulation of African Easterly Waves in a Global Climate Model

Abstract African easterly waves (AEWs) exert significant influence on local and downstream high-impact weather including tropical cyclone (TC) genesis over the Atlantic. Accurate representation of AEWs in climate and weather prediction models therefore is necessary for skillful predictions. In this study, we examine simulated AEWs, including their evolution, vertical structure, and linkage to tropical cyclone genesis, in the NASA Goddard Earth Observing System Model, version 5 (GEOS-5) atmospheric global climate model. Identified by the leading empirical orthogonal function mode of time-filtered precipitation, the observed westward propagating AEWs along the southern track over the Atlantic are largely captured in GEOS-5, but with a slower phase speed and significantly weaker amplitude downstream off the West Africa coast. The weak downstream development of AEWs in GEOS-5 is accompanied by much reduced TC genesis over the main development region. Further analyses suggest that the slow westward propagation and weaker AEW amplitude downstream can be ascribed to a weak African easterly jet, while overestimated negative (positive) meridional potential vorticity (PV) gradients appear to the north (south) of 10°N in GEOS-5. The greatly overestimated positive meridional PV gradient to the south of 10°N is expected to generate strong horizontal stretching in the AEW wave pattern in the model, which hinders organization of convection and its feedback to sustain the AEW development. Persistent and vigorous AEW precipitation in the Guinea Highlands of the West Africa coast could also be responsible for reduced westward propagation of AEWs in the model.

Improved advection, resolution, performance, and community access in the new generation (version 13) of the high-performance GEOS-Chem global atmospheric chemistry model (GCHP)

Abstract. We describe a new generation of the high-performance GEOS-Chem (GCHP) global model of atmospheric composition developed as part of the GEOS-Chem version 13 series. GEOS-Chem is an open-source grid-independent model that can be used online within a meteorological simulation or offline using archived meteorological data. GCHP is an offline implementation of GEOS-Chem driven by NASA Goddard Earth Observing System (GEOS) meteorological data for massively parallel simulations. Version 13 offers major advances in GCHP for ease of use, computational performance, versatility, resolution, and accuracy. Specific improvements include (i) stretched-grid capability for higher resolution in user-selected regions, (ii) more accurate transport with new native cubed-sphere GEOS meteorological archives including air mass fluxes at hourly temporal resolution with spatial resolution up to C720 (∼ 12 km), (iii) easier build with a build system generator (CMake) and a package manager (Spack), (iv) software containers to enable immediate model download and configuration on local computing clusters, (v) better parallelization to enable simulation on thousands of cores, and (vi) multi-node cloud capability. The C720 data are now part of the operational GEOS forward processing (GEOS-FP) output stream, and a C180 (∼ 50 km) consistent archive for 1998–present is now being generated as part of a new GEOS-IT data stream. Both of these data streams are continuously being archived by the GEOS-Chem Support Team for access by GCHP users. Directly using horizontal air mass fluxes rather than inferring from wind data significantly reduces global mean error in calculated surface pressure and vertical advection. A technical performance demonstration at C720 illustrates an attribute of high resolution with population-weighted tropospheric NO2 columns nearly twice those at a common resolution of 2∘ × 2.5∘.

Change in Tropospheric Ozone in the Recent Decades and Its Contribution to Global Total Ozone

AbstractTropospheric ozone is a key chemically active trace gas and radiative forcer. Understanding its long‐term changes is important to properly interpret observed changes in total column ozone and stratospheric ozone recovery. We investigate global and regional tropospheric ozone changes and their impact on total column ozone during 2005–2018 using satellite measurements and the NASA Goddard Earth Observing System Chemistry Climate Model (GEOSCCM). Global total ozone increased ∼4 DU during 2005–2018 (+0.28 ± 0.06 DU yr−1) as inferred from Ozone Monitoring Instrument (OMI). Consistent with previous studies, the OMI/MLS (Microwave Limb Sounder) derived global tropospheric ozone increased 2.2 DU during this period, 60% of the global total column ozone increase. While GEOSCCM reproduces reasonably well the total column increase, it overestimates the stratospheric ozone increase and underestimates the tropospheric ozone increase. We find that the tropospheric ozone increases are likely attributed to a growth of regional emissions of key ozone precursors, especially volatile organic compounds as reflected by the positive trends in formaldehyde (CH2O). Although carbon monoxide (CO) has been decreasing everywhere around the globe, it has relatively small impact on the tropospheric ozone trend. Trends in nitrogen dioxide (NO2) vary with regions, and these changes counteract or reinforce the positive effects of CH2O on the tropospheric ozone increases. The model underestimates the observed tropospheric ozone increase, especially over the US and Europe, because of underestimated NO2 emissions changes used in the model. The stratospheric ozone contribution increases during this period in the Northern Hemisphere and contributes to the tropospheric ozone increase.

High-resolution air quality forecasts for LMIC using a combination of machine learning and earth model simulations

Background and aim: With the recent advancement of Earth observation, computer simulations, and low-cost monitoring technologies, the capacity to obtain accurate geospatial data has increased tremendously, particularly for locations without expansive in-situ monitoring networks. Here we present an optimized machine learning model to estimate near real-time air pollutant concentrations in selected locations in Africa and Latin America using a combination of NASA Goddard Earth Observing System Composition Forecasts (GEOS-CF) and low-cost sensor data. Methods: We use a machine learning approach to estimate near real-time air pollutants concentration at selected locations across Africa and Latin America. Several meteorological and chemical parameters are retrieved from NASA’s GEOS-CF to train a bias corrector model and predict corrected concentration estimates for these locations which are then validated against local monitoring data. We also conduct an explainability approach via SHAP Analysis to quantify the model contributing factors across these locations and track the model performance in extreme conditions. Results: The optimized machine learning model shows good agreement with ground air quality data, with R2 values from 0.61- 0.65 for locations in Mexico City (Mexico), Bogotá (Columbia) and Kigali (Rwanda). Via SHAP analysis, we demonstrate the need to conduct a measure of variance to determine model performance for different conditions and intervals, knowing that the model performance can be affected by various training conditions. Conclusion: Combining observations and optimized model simulations using machine learning techniques can significantly improve air quality forecasts in low- and middle-income countries, where the rapid pace of industrialization and communities are highly susceptible to air pollution health effects, and rarely have local air quality data and health risks alerting systems in place. These results are being used to assist air quality managers and environmental agencies in these locations to improve risk communication and reduce health burdens associated with outdoor air pollution.

Impact of Assimilating Adaptively Thinned AIRS Cloud-Cleared Radiances on the Analysis of Polar Lows and Mediterranean Sea Tropical-Like Cyclone in a Global Modeling and Data Assimilation Framework

Abstract Polar lows, and mesoscale convective cyclones bearing resemblance to tropical cyclones but originating outside of the tropics, are storms that are challenging to represent accurately in global analyses and models because of their small size, rapid growth at subsynoptic scales, occurrence in data poor oceanic regions, and difficulties in objectively validating them in analysis. Building on previous positive results obtained with respect to the representation of tropical cyclones (TCs) in a global model, a set of observing system experiments (OSEs) performed using the NASA Goddard Earth Observing System (GEOS, version 5) are investigated, focusing on three case studies—a polar low in the Sea of Okhotsk, a polar low in the Southern Ocean, and a Mediterranean Sea tropical-like cyclone that occurred during the boreal fall season of 2014. Experiments assimilating adaptively thinned cloud-cleared hyperspectral infrared radiances from the Atmospheric Infrared Sounder (AIRS) instrument on board the NASA Aqua satellite, with higher density in the vicinity of each storm and its pre-cyclogenesis environment, and lower density elsewhere, demonstrate a positive impact on the analyzed representation of each storm. The adaptive thinning experiments improve the storm intensity and structure, including vertical alignment, depth, symmetry, strength, and compactness of warm core compared to the reference experiments. The results suggest that jet-level processes associated with extremely strong horizontal velocity gradients as represented in the model analysis can be useful to locate dynamically active regions of the extratropical atmosphere where denser data coverage is likely to improve the analyzed representation of polar lows and other similar marine mesoscale convective cyclones. Significance Statement Extratropical maritime mesoscale convective cyclones are short-lived, elusive features that are difficult to represent accurately in global analyses. Previous work by this team demonstrated a positive impact of an adaptive thinning methodology for infrared radiances applied to the tropical cyclone (TC) analysis. The methodology allows a relatively greater volume of radiance data to be assimilated around TCs within a TC-centered moving domain in a global model, yielding an improvement in TC structure and intensity forecast. A similar approach is explored here for two polar lows and a Mediterranean Sea tropical-like cyclone, wherein infrared radiances are more densely assimilated in the vicinity of each storm and its pre-cyclogenesis environment, resulting in a positive impact on the representation of the storm. Strong jet-level horizontal velocity gradients appear to precede each storm, and could be used to automate the adaptive thinning strategy in the future.

NASA GEOS Composition Forecast Modeling System GEOS-CF v1.0: Stratospheric Composition.

The NASA Goddard Earth Observing System (GEOS) Composition Forecast (GEOS‐CF) provides recent estimates and 5‐day forecasts of atmospheric composition to the public in near‐real time. To do this, the GEOS Earth system model is coupled with the GEOS‐Chem tropospheric‐stratospheric unified chemistry extension (UCX) to represent composition from the surface to the top of the GEOS atmosphere (0.01 hPa). The GEOS‐CF system is described, including updates made to the GEOS‐Chem UCX mechanism within GEOS‐CF for improved representation of stratospheric chemistry. Comparisons are made against balloon, lidar, and satellite observations for stratospheric composition, including measurements of ozone (O3) and important nitrogen and chlorine species related to stratospheric O3 recovery. The GEOS‐CF nudges the stratospheric O3 toward the GEOS Forward Processing (GEOS FP) assimilated O3 product; as a result the stratospheric O3 in the GEOS‐CF historical estimate agrees well with observations. During abnormal dynamical and chemical environments such as the 2020 polar vortexes, the GEOS‐CF O3 forecasts are more realistic than GEOS FP O3 forecasts because of the inclusion of the complex GEOS‐Chem UCX stratospheric chemistry. Overall, the spatial patterns of the GEOS‐CF simulated concentrations of stratospheric composition agree well with satellite observations. However, there are notable biases—such as low NO x and HNO3 in the polar regions and generally low HCl throughout the stratosphere—and future improvements to the chemistry mechanism and emissions are discussed. GEOS‐CF is a new tool for the research community and instrument teams observing trace gases in the stratosphere and troposphere, providing near‐real‐time three‐dimensional gridded information on atmospheric composition.

Using an Explainable Machine Learning Approach to Characterize Earth System Model Errors: Application of SHAP Analysis to Modeling Lightning Flash Occurrence

AbstractComputational models of the Earth System are critical tools for modern scientific inquiry. Efforts toward evaluating and improving errors in representations of physical and chemical processes in these large computational systems are commonly stymied by highly nonlinear and complex error behavior. Recent work has shown that these errors can be effectively predicted using modern Artificial Intelligence (A.I.) techniques. In this work, we go beyond these previous studies to apply an explainable A.I. technique to not only predict model errors but also move toward understanding the underlying reasons for successful error prediction. We use XGBoost classification trees and SHapley Additive exPlanations analysis to explore the errors in the prediction of lightning occurrence in the NASA Goddard Earth Observing System model, a widely used Earth System Model. This explainable error prediction system can effectively predict the model error and indicates that the errors are strongly related to convective processes and the characteristics of the land surface.

Sensitivity of low‐tropospheric Arctic temperatures to assimilation of AIRS cloud‐cleared radiances: Impact on midlatitude waves

AbstractIn this work it is shown that the prediction of individual midlatitude waves in a global forecast framework may benefit from the assimilation of cloud‐cleared radiances (CCRs) from the Atmospheric Infrared Sounder (AIRS). The assimilation of CCRs, relative to clear‐sky radiances, provides more information in high‐latitude areas affected by broken low‐level stratus clouds, which are common over north polar latitudes during the summer and autumn seasons. This added information in cloudy regions produces slightly lower low‐level temperatures and lower mid‐tropospheric heights (through hydrostatic adjustment) over the Arctic and part of northeastern Siberia. In areas that are both dynamically unstable (with the potential for rapid error growth) and data void (as observed by clear‐sky radiances), the assimilation of CCRs provides valuable information that can improve the representation of individual baroclinic waves and their subsequent forecasts. The assimilation of AIRS CCRs slightly modifies the geopotential height gradients between the Arctic and the midlatitudes, leading to improvements in the forecast of individual baroclinic waves, as is shown through three case‐studies. The set of observing system experiments (OSEs) is performed with the NASA Goddard Earth Observing System (GEOS) data assimilation and forecast system during boreal autumn 2014. AIRS CCRs are thinned to approximately one quarter the density of operationally assimilated AIRS radiances, consistent with their higher information content. Global 6‐hourly analyses are produced from 01 September to 10 November 2014 and seven‐day forecasts are initialized at 0000 UTC daily. Since the CCR methodology is widely applicable, these findings are also relevant to other infrared sensors.

The response of the Amazon ecosystem to the photosynthetically active radiation fields: integrating impacts of biomass burning aerosol and clouds in the NASA GEOS Earth system model

Abstract. The Amazon experiences fires every year, and the resulting biomass burning aerosols, together with cloud particles, influence the penetration of sunlight through the atmosphere, increasing the ratio of diffuse to direct photosynthetically active radiation (PAR) reaching the vegetation canopy and thereby potentially increasing ecosystem productivity. In this study, we use the NASA Goddard Earth Observing System (GEOS) model with coupled aerosol, cloud, radiation, and ecosystem modules to investigate the impact of Amazon biomass burning aerosols on ecosystem productivity, as well as the role of the Amazon's clouds in tempering this impact. The study focuses on a 7-year period (2010–2016) during which the Amazon experienced a variety of dynamic environments (e.g., La Niña, normal years, and El Niño). The direct radiative impact of biomass burning aerosols on ecosystem productivity – called here the aerosol diffuse radiation fertilization effect – is found to increase Amazonian gross primary production (GPP) by 2.6 % via a 3.8 % increase in diffuse PAR (DFPAR) despite a 5.4 % decrease in direct PAR (DRPAR) on multiyear average during burning seasons. On a monthly basis, this increase in GPP can be as large as 9.9 % (occurring in August 2010). Consequently, the net primary production (NPP) in the Amazon is increased by 1.5 %, or ∼92 Tg C yr−1 – equivalent to ∼37 % of the average carbon lost due to Amazon fires over the 7 years considered. Clouds, however, strongly regulate the effectiveness of the aerosol diffuse radiation fertilization effect. The efficiency of this fertilization effect is the highest in cloud-free conditions and linearly decreases with increasing cloud amount until the cloud fraction reaches ∼0.8, at which point the aerosol-influenced light changes from being a stimulator to an inhibitor of plant growth. Nevertheless, interannual changes in the overall strength of the aerosol diffuse radiation fertilization effect are primarily controlled by the large interannual changes in biomass burning aerosols rather than by changes in cloudiness during the studied period.

Impacts of the Eruption of Mount Pinatubo on Surface Temperatures and Precipitation Forecasts With the NASA GEOS Subseasonal‐to‐Seasonal System

AbstractA contemporary seasonal forecasting system is used to study the impacts of a volcanic sulfate injection into the stratosphere on the seasonal forecasts for surface temperatures, the El Niño Southern Oscillation (ENSO), and precipitation. The focus is a case study of the June 1991 eruption of Mt. Pinatubo, Philippines and the period from July 1991 to February 1992. Version 2 of the Goddard Earth Observing System (GEOS) subseasonal‐to‐seasonal (S2S) forecasting system is used in this study. GEOS‐S2S includes the GOddard Chemistry, Aerosols, Radiation and Transport (GOCART) aerosol module, which allows to prognostically simulate aerosol distributions. GOCART is coupled to the radiation and cloud modules to include the impact of the eruption on forecasted radiation and precipitation. The coupled GEOS‐S2S system was initialized in May 1991 with fields based on observations to produce ten‐member 9‐month forecasts with and without the volcanic sulfur injection. The results of these ensemble experiments demonstrate that including Mt. Pinatubo in seasonal forecasts would improve the forecasts of the 1991–1992 global mean temperature and precipitation but worsen the forecast of ENSO by strengthening of El Niño beyond what showed in observations. Most significant changes in the forecasts of temperatures and precipitation are limited to the tropics. The only land area where the inclusion of Pinatubo significantly lowered the forecasted precipitation is tropical Africa.

Contribution of Meteorological Downscaling to Skill and Precision of Seasonal Drought Forecasts

AbstractResearch in meteorological prediction on sub-seasonal to seasonal (S2S) timescales has seen growth in recent years. Concurrent with this, demand for seasonal drought forecasting has risen. While there is obvious synergy between these fields, S2S meteorological forecasting has typically focused on low resolution global models, while the development of drought can be sensitive to the local expression of weather anomalies and their interaction with local surface properties and processes. This suggests that downscaling might play an important role in the application of meteorological S2S forecasts to skillful forecasting of drought. Here, we apply the Generalized Analog Regression Downscaling (GARD) algorithm to downscale meteorological hindcasts from the NASA Goddard Earth Observing System (GEOS) global S2S forecast system. Downscaled meteorological fields are then applied to drive offline simulations with the Catchment Land Surface Model (CLSM) to forecast United States Drought Monitor (USDM) style drought indicators derived from simulated surface hydrology variables. We compare the representation of drought in these downscaled hindcasts to hindcasts that are not downscaled, using the North American Land Data Assimilation System Phase 2 (NLDAS-2) dataset as an observational reference. We find that downscaling using GARD improves hindcasts of temperature and temperature anomalies, but the results for precipitation are mixed and generally small. Overall, GARD downscaling led to improved hindcast skill for total drought across the Contiguous United States (CONUS), and improvements were greatest for extreme (D3) and exceptional (D4) drought categories.

The Response of the Amazon Ecosystem to the Photosynthetically Active Radiation Fields: Integrating Impacts of Biomass Burning Aerosol and Clouds in the NASA GEOS ESM

Abstract. The Amazon experiences fires every year, and the resulting biomass burning aerosols, together with cloud particles, influence the penetration of sunlight through the atmosphere, increasing the ratio of diffuse to direct photosynthetically active radiation (PAR) reaching the vegetation canopy and thereby potentially increasing ecosystem productivity. In this study, we use the NASA Goddard Earth Observing System (GEOS) model running with coupled aerosol, cloud, radiation, and ecosystem modules to investigate the impact of Amazon biomass burning aerosols on ecosystem productivity, as well as the role of the Amazon’s clouds in tempering the impact. The study focuses on a seven-year period (2010–2016) during which the Amazon experienced a variety of dynamic environments (e.g., La Nina, normal years, and El Nino). The radiative impacts of biomass burning aerosols on ecosystem productivity – call here the aerosol light fertilizer effect – are found to increase Amazonian Gross Primary Production (GPP) by 2.6 % via a 3.8 % increase in diffuse PAR (DFPAR) despite a 5.4 % decrease in direct PAR (DRPAR) on multiyear average. On a monthly basis, this increase in GPP can be as large as 9.9 % (occurring in August 2010). Consequently, the net primary production (NPP) in the Amazon is increased by 1.5 %, or ~92 TgCyr−1– equivalent to ~37 % of the carbon lost due to Amazon fires over the seven years considered. Clouds, however, strongly regulate the effectiveness of the aerosol light fertilizer effect. The efficiency of the fertilizer effect is highest for cloud-free conditions and linearly decreases with increasing cloud amount until the cloud fraction reaches ~0.8, at which point the aerosol-influenced light changes from being a stimulator to an inhibitor of plant growth. Nevertheless, interannual changes in the overall strength of the aerosol light fertilizer effect are primarily controlled by the large interannual changes in biomass burning aerosols rather than by changes in cloudiness during the studied period.