Impacts of aerosol–monsoon interaction on rainfall and circulation over Northern India and the Himalaya Foothills
- 2.5k Downloads
The boreal summer of 2008 was unusual for the Indian monsoon, featuring exceptional heavy loading of dust aerosols over the Arabian Sea and northern-central India, near normal all-India rainfall, but excessive heavy rain, causing disastrous flooding in the Northern Indian Himalaya Foothills (NIHF) regions, accompanied by persistent drought conditions in central and southern India. Using the NASA Unified-physics Weather Research Forecast (NUWRF) model with fully interactive aerosol physics and dynamics, we carried out three sets of 7-day ensemble model forecast experiments: (1) control with no aerosol, (2) aerosol radiative effect only and (3) aerosol radiative and aerosol-cloud-microphysics effects, to study the impacts of aerosol-monsoon interactions on monsoon variability over the NIHF during the summer of 2008. Results show that aerosol-radiation interaction (ARI), i.e., dust aerosol transport, and dynamical feedback processes induced by aerosol-radiative heating, plays a key role in altering the large-scale monsoon circulation system, reflected by an increased north-south tropospheric temperature gradient, a northward shift of heavy monsoon rainfall, advancing the monsoon onset by 1–5 days over the HF, consistent with the EHP hypothesis (Lau et al. in Clim Dyn 26(7–8):855–864, 2006). Additionally, we found that dust aerosols, via the semi-direct effect, increase atmospheric stability, and cause the dissipation of a developing monsoon onset cyclone over northeastern India/northern Bay of Bengal. Eventually, in a matter of several days, ARI transforms the developing monsoon cyclone into meso-scale convective cells along the HF slopes. Aerosol-Cloud-microphysics Interaction (ACI) further enhances the ARI effect in invigorating the deep convection cells and speeding up the transformation processes. Results indicate that even in short-term (up to weekly) numerical forecasting of monsoon circulation and rainfall, effects of aerosol-monsoon interaction can be substantial and cannot be ignored.
KeywordsMonsoon Rainfall Lower Troposphere Deep Convection Indian Monsoon Monsoon Onset
The Northern India and Himalayan foothills (NIHF) region is an essential component of the Indian monsoon (Gadgil et al. 2003). Summer monsoon rainfall over this region feeds the Ganges, and the Indus rivers, providing fresh water essential for agriculture, industrial and livelihood over hundreds of millions people in the region. The Himalayan foothills (HF), an escarpment that rises to over 5 km above the Indo-Gangetic Plain (IGP) over Northern India (NI), provides a major barrier to the northward advance of the Indian monsoon rainfall, as well as contributing to the thermal contrast between the Tibetan Plateau and the Indian Ocean (Boos and Kuang 2010; Wu et al. 2012). It also provides strong orographic forcing to the prevailing southwesterly monsoon flow, giving rise to strong vertical ascent and development of vigorous thunderstorm cells with intense precipitation during the monsoon season (Houze et al. 2007; Rasmussen and Houze 2012; Das et al. 2014). The importance of the NIHF orography in forcing strong convection and extreme precipitations, in relationship to the large-scale Indian monsoon has also been reported (Barros and Lang 2003; Barros and Lattenmaier 1994).
Recently, an increasing number of studies have suggested that absorbing aerosols (mainly desert dusts and carbonaceous aerosols) can affect the interannual and intraseasonal variability of the Indian monsoon rainfall (Lau et al. 2006; Lau and Kim 2006; Lau 2014; Manoj et al. 2011; Hazra et al. 2013; Vinoj et al. 2014; Kim et al. 2015; D’Errico et al. 2015; Sanap and Pandithurai 2015, and many others). Based on these studies, Lau (2016) has proposed a new paradigm arguing that natural aerosol, particularly dust and carbonaceous aerosols from natural sources should be considered an essential component of an intrinsic aerosol-monsoon weather and climate system. Specifically for the NIHF region, the Elevated Heat Pump (EHP) hypothesis (Lau et al. 2006) posited that through atmospheric diabatic heating and circulation feedbacks, atmospheric heating due to absorbing aerosols from both anthropogenic and natural sources accumulated over the Indo-Gangetic Plain during the pre-monsoon period can, lead to increased rainfall and advance of the monsoon rainy season over the NIHF. At present, effects of absorbing aerosols in increasing monsoon rainfall and circulation consistent with EHP, but with significantly different regional spatial and temporal details have been found in many global climate models (Lau et al. 2006; Meehl et al. 2008; Randles and Ramaswamy 2008; Wang et al. 2009; Collier and Zhang 2009; Henriksson et al. 2014; D’Errico et al. 2015; Jin et al. 2015). One of the main reasons for the regional differences is that state-of-the-art climate models, due to their coarse resolution generally, are still unable to resolve the complex topography and interactions of monsoon winds, precipitation and aerosol transport, sources and sink processes over the NIHF. For more realistic simulations, the use of high-resolution regional atmospheric model with realistic representation of aerosol and monsoon processes is required. In this paper, using the NASA Unified Physics Weather Research Forecast (NU-WRF) model, we have carried out numerical experiments aimed at providing a better understanding of the physical processes involved in aerosol-monsoon interactions, including aerosol radiative and microphysical effects, as well as induced dynamical feedbacks over the complex topography of the NIHF region. The experiments will be conducted based on a case study of the 2008 Indian monsoon.
1.1 The 2008 Indian monsoon
2 Modeling strategy and methodology
We use the NASA Unified physics Weather Research Forecast (NUWRF) regional atmospheric model to carry out ensemble forecast experiments to investigate the role of aerosol-monsoon dynamic interaction in the NIHF region. NUWRF has been developed at GSFC to unify physics in WRF with NASA’s current physics package for aerosol and cloud microphysics, as well as assimilation systems, common to GEOS-5, the Goddard Cumulus Ensemble cloud model (GCE) (Tao et al. 2009), the Goddard Chemistry and Radiation Transport (GOCART) model (Chin et al. 2004, 2009, 2014; Ginoux et al. 2001), and the Land data Information System (LIS) (Peters-Lidard et al. 2007, 2015; Tao et al. 2013; Lang et al. 2014; Shi et al. 2014). Briefly, GOCART uses the following modules in aerosol simulation: emission, which includes dust, sulfur, black carbon, organic carbon, and sea-salt emissions; chemistry, which includes in-air and in-cloud oxidations of sulfate precursors (SO2 and DMS); turbulent dry deposition, which is calculated by aerodynamic resistance; gravitational settling, which is a function of air viscosity and particle size; and wet deposition, which accounts for the scavenging of soluble species in convective updrafts and rainout/washout in large-scale precipitation.
Dust size distribution is calculated by solving the continuity equation for eight size bins ranging from 0.1 to 10 μm in radii, representing mineral dust types from fine clay to coarse sand particles. The continuity equation includes macroscopic advection by winds, parameterized eddy diffusion and moist convection. The removal mechanisms include dry deposition at the surface by impaction, wet deposition in and below clouds, and gravitational settling. Dust emission is computed as a function of the surface wind at 10 m, a threshold frictional velocity, and a source function which is depended on geographic locations and dust sizes (Ginoux et al. 2001; Gillette and Passi 1988).
For aerosol-cloud microphysics, NUWRF computes cloud droplet concentration using aerosol mass directly predicted by GOCART/WRF-Chem at each time step. For a given air temperature and supersaturation, cloud condensation nuclei (CCN) is calculated from the aerosol species predicted by GOCART based on the Koehler curve (Koehler et al. 2006; Andreae and Rosenfeld 2008), while the concentration of ice nuclei (IN) is obtained following the approach of Demott et al. (2010). Both CCN and IN are diagnostic parameters calculated from the WRF-Chem/GOCART-predicted aerosol mass concentrations in the Goddard one-moment microphysics scheme. CCN is used to calculate the auto-conversion of cloud liquid water (Qc) to rain water (Qr) following Liu and Daum (2004), while IN is used to calculate 1) the conversion of cloud ice (Qi) to snow (Qs) due to the Bergeron process and 2) the growth of cloud liquid water (Qc) to cloud ice (Qi) due to deposition.
The NUWRF team has also coupled GOCART with LIS which provides key information (e.g., soil moisture and soil porosity) that GOCART uses to calculate dust emissions. GOCART has also been coupled with the Goddard radiation and microphysics scheme to simulate the direct and indirect aerosol effects on energy budget, cloud, and precipitation. Parameterization schemes to estimate biogenic secondary organic aerosols have also been integrated from WRF/Chem into NUWRF. In addition, several model utilities have been developed and modified to facilitate NUWRF applications, including implementation of time varying GOCART aerosols types and distribution as initial and boundary conditions, and incorporation of pollutant emissions inventories for both anthropogenic and natural sources. For the present experiments, the single scattering albedo of dust aerosol at the visible band (0.4–0.7 μm) were specified in the range 0.95 to 0.79, representing 8 size categories of dust from fine clay to coarse dust particles. The much higher absorption (lower single scattering albedo) for larger size particles is consistent with recent observations indicating coating of fine black carbon particle on the larger size sand particles in Asian monsoon regions (Ganguly et al. 2005; Satheesh et al. 2008; Eck et al. 2010).
For our model experiments, we followed the ensemble forecast methodology commonly used in monsoon numerical weather prediction up to a week to investigate the impacts of aerosol-monsoon dynamic interaction on short time scales. We used nested domains with 9 km resolution in an inner domain (60°–100°E, 10°–40°N) covering the Indian subcontinent and eastern Indo-China, and an outer domain with 27 km resolution, spanning the large-scale monsoon domain (40–120°E, 5°S–40°N), with 61 layers in the vertical (Fig. S1). We selected the early monsoon period June 1–July 15 2008 for our study. This period covered the monsoon onset transition phase (June 11–17), and several subsequent heavy rain events over NIHF (See discussion in Sect. 3.2). We carried out 45, 7-day forecast experiments forced by prescribed observed sea surface temperature, and with atmospheric and aerosol boundary conditions initialized at intervals of one day starting from June 1, first with no aerosol (NA) effects, which is still the common practice for most numerical weather forecasts in monsoon regions (Krishnamurti et al. 1991). A second identical set of experiments was carried out but with the inclusion of aerosol radiative (RAD) effect. A third set of experiments same as the control, except with both aerosol radiative and microphysics interactive effects (RADM) were carried out for the onset period June 11–17. Initial and boundary conditions for meteorology and aerosols were obtained respectively from NCEP GFS 1° × 1° global analysis, and from GOCART. By comparing the RAD and RADM to NA forecasts over the NIHF region in a realistic setting with a high-resolution regional model and sophisticated aerosol physics, we can gain a better understanding of intrinsic, aerosol-monsoon interaction for the Indian monsoon, involving no change in anthropogenic aerosol emissions. In following discussions, all model anomaly quantities are defined with respect to NA. For brevity, we shall refer to RAD anomalies as due to Aerosol-Radiative Interaction (ARI) effects. Effects of Aerosol-Cloud-microphysics-Interaction (ACI) will be diagnosed from comparison between RADM and RAD anomalies. Aerosol impacts will be examined using two different kinds of averaging. First, ensemble forecast averaging for day-1 through day-7 for a specific date, will be compared to observation to assess aerosol impacts on forecast skills. Second, ensemble averages of all forecasts as a function of forecast days, will be used to assess the aerosol impact as a function of forecast lead time. For comparing model results to observations, we use rainfall data from the Tropical Precipitation Measuring Mission (TRMM), and winds and temperature data from the NASA MERRA2, Modern Era Retrospective analysis for Research and Applications -version 2 (Bosilovich et al. 2015).
To facilitate comparison of model results to observations, and set the stage for the model output analysis, a brief description of the observed large-scale features of the Indian monsoon of 2008 is presented first.
3.1 Observed large-scale monsoon features during June 2008
3.2 Impact on model forecasts
The model mean rainfall and wind patterns show broad similarities to observations (Fig. 3c, d), but with notable differences. The model low-level westerlies are too strong. As a result, precipitation upwind over the mountain ridges of the Western Ghats, and Indo-China are overestimated. The model also missed the heavy rainfall band extending from the east coast of northern India into the Bay of Bengal. These are mostly due to the excessive monsoon low-level westerlies across central and southern India. As shown in following discussion, these biases are not related to model aerosol physics, because they are essential unchanged in the RAD and RADM simulations. Another notable feature is the excessive model rainfall over the southern Indo-Gangetic Plain, and deficient rainfall over the HF compared to TRMM. In the following, we will demonstrate how this bias is mitigated in the RAD and RADM experiments and how we can use the incremental improvement to explore the physical processes underlying aerosol-monsoon interactions over NIHF region. Despite the aforementioned discrepancies, the overall AOD and rainfall distributions in the model are consistent with dust aerosol emissions from the arid region of Southwest Asia, the Thar desert, and Middle East deserts and transport across the Arabia Sea, with strong wet removal in rainy regions.
3.3 Large-scale interactions
The ARI effects on precipitation and circulation are strongest over the domain covering eastern India and the western Bay of Bengal (80°–90°E). As shown in Fig. 8e, the ARI-induced upper tropospheric warming over the Tibetan Plateau and regions to the south are very pronounced. This occurs in conjunction with a pronounced elevated thick aerosol layer, pushing against the southern slopes of the Tibetan Plateau by the increasing monsoon southerlies (See Fig. 6b). Here the dust loading is lower compared to that over the Arabian Sea and western India, due to removal by both dry and wet depositions during the long-distance transport from the Middle East deserts to eastern India. However, still noticeable are the cooling near the top of the dust layer, warming within and below the dust layer and near the surface, characteristic of light-absorbing dust layers (Zhu et al. 2007; Lau et al. 2009). A signal of the aerosol semi-direct effect reducing convective potential, in the form of strong surface cooling due to aerosol solar attenuation and warming induced by aerosol heating can be found over the HF (25–35°N, Fig. 8e). However, this effect is overpowered by the increased low-level moisture advection into this region by the strengthened southwesterlies, as evident in the increased moist static energy in the lower troposphere (700 hPa) relative to the regions above (See discussion in Sect. 3.4, Fig. 11). Except within the dust layer, specific humidity is substantially increased over the entire atmospheric column. The increase is most pronounced and coincides with a strong low-level westerly jet near 25–30N, (Fig. 8f). Coupled to the wind, temperature, moisture and aerosol anomalies, is an enhanced local meridional circulation, with strong rising motion in the HF, and sinking motion to the south (Fig. 8g), and the northward displacement of the maximum rainfall over the HF (Fig. 8h).
3.4 Mesoscale Interactions
The invigoration of deep convection by ARI and ACI in the HF region is further examined via the vertical distributions of hydrometeors associated with the deepening convection in the ND and SD. Without aerosol effects (NA), the distributions of liquid water (cloud and rain), cloud ice, snow and graupel indicate stronger development of deep convection in SD than in ND (Fig. 10a, b). Comparing RAD (Fig. 10c, d) with NA (Fig. 10a, b), ARI shifts the maximum cloud ice upwards from 250 to 200 hPa, enhances the concentration of snow and graupel in the 500–300 hPa layer, and increases liquid water content in the lower troposphere in the ND. In the SD, the effects are reversed, with substantial reduction in deep convection as evidenced in substantial loss of cloud ice, and snow and graupel in the mid to upper troposphere, as well as suppressed cloud and rain water in the lower troposphere. Comparing RADM (Fig. 10e, f) and RAD (Fig. 10c, d) anomalies indicates that deep convection is further enhanced by ACI in ND, as evident in the larger magnitude and extent of the increase in cloud ice, and snow/graupel in the mid- and upper troposphere. This appears to be at the expense of reduced rate of increase in cloud and rain water in the lower troposphere compared to RAD. In the SD, the stronger reduction hydrometeors of all kinds suggest a further suppression of deep convection by ACI. For the domain as a whole, the change in total rainfall due to inclusion of ACI is relatively small (<5%). Most of the changes by ACI appear to be in the nature of the deep convection and not in the total rain. These results are consistent with modeling and observations indicating enhanced glaciation, and formation of deeper clouds by aerosols in a moist environment [Rosenfeld et al. 2008; Fan et al. 2013].
The changes in the nature of the deep convection induced by ARI and ACI can be viewed in terms of changes in moist static energy (MSE) of the atmosphere over the ND and SD during the onset phase (Fig. 11). Judging from the stronger negative gradient (reduction with height) in mean MSE in the lower troposphere (Fig. 11a, b), it can be inferred that in the absence of aerosol influence (NA), the SD is more (compared to the ND) convectively unstable, and convection is likely to be deeper, as reflected in the larger loading of hydrometeors of all types (Fig. 10a, b). Under ARI forcing, the lower troposphere in ND (Fig. 11c, d) is further destabilized with increase in negative gradient MSE between 800 hPa and 600 hPa, while the mid- to upper troposphere between 600 and 300 hPa are stabilized, as evidence in the positive gradients in MSE between these two levels. Above 300 hPa, the MSE reflects again a convective de-stabilization of the atmosphere. Computations of separate contributions to the MSE (Fig. S3) shows that the destabilization in the lower troposphere is mainly due to increased moisture, while stabilization is mainly due to temperature change, i.e., aerosol semi-direct effect in heating the lower troposphere and cooling of the surface. Changes in MSE in the mid-to upper troposphere is associated with temperature change due to moist adiabatic ascent of heated air from below reaching maximum altitude. Essentially, in the ND, the low-level moisture effect overpowers the aerosol semi-effect, invigorating more intense convection and rainfall. Clearly, ACI enhances the MSE convective stability modulation due to ARI, favoring enhanced deep convection in ND (Fig. 11e). In contrast, for the SD under ARI (Fig. 11f), both the aerosol semi-direct effect and the reduction in moisture by advection (see discussion related to Fig. 7) stabilizes the lower troposphere, leading to the dissipation of the monsoon depression and suppressed precipitation. This effect is further amplified by ACI (Fig. 11f). As noted previously, the northward advection of the un-precipitated moisture from suppressed rainfall in SD by the strengthened southerlies can further increase in convective instability and rainfall in ND. This is the essence of the EHP effect. Overall, it can be argued that the MSE changes provide the physical underpinnings of anomalous deep convection, rainfall and moisture transport induced by ARI and ACI in the NIHF region.
The complex and steep topography of the Himalaya foothills facilitate the build up of thick layers of dust aerosols transported by monsoon southwesterlies from the Arabian Peninsular deserts across the Arabian Sea, and accumulated over northwestern and central India during the pre- and early monsoon season of 2008.
Via aerosol induced dynamical feedback, ARI warms the upper troposphere over the Tibetan Plateau, increases the mid-to-upper tropospheric north-south temperature gradient, and thus strengthens the early Indian monsoon by shifting maximum monsoon rainfall from the Indo-Gangetic Plain to the Himalayan foothills regions, and advancing the timing of monsoon onset over HF by 1–5 days.
Aerosols have both local and non-local effects, which interacts with each other. Generally, regions of high aerosol loading are associated with drier air masses and suppressed rainfall, and heavy rain regions with low aerosol loading due to wash-out. The semi-direct (local) effect of dust aerosols leads to the stabilization and weakening of a developing monsoon depression over northeastern India and northern Bay of Bengal during the onset phase of the monsoon.
Non-local effects are exerted mainly through moisture and circulation feedback. Heating by thick dust layer over the Arabian Sea and northern India, enhances the large-scale southwesterly low-level monsoon flow and transport of moisture from the Arabian Sea to the Indian subcontinent. The enhanced southwesterly moisture transport increases low-level MSE, leading to the development of intense convective cells over the Himalaya Foothill regions.
While ARI dominates the large-scale aerosol-monsoon interaction, ACI further enhances the ARI effects, by suppressing development of monsoon depression over northeastern India, while intensifying the ice-phase precipitation and deep cloud processes in the HF, and thereby speeding up the transformation the monsoon depression into intense meso-scale cells with heavy rain in the forms line convection along the rugged terrains of the HF.
Our results are in general agreement with recent modeling studies using the WRF-CHEM. Krishnamurti et al. (2013) showed that the development of monsoon depression is suppressed by aerosol via increased cloud condensation nuclei (CNN), over northeastern India and Bay of Bengal. However, their results did not indicate development of mesoscale-scale deep convection over the Himalayan foothills, most likely because aerosol number concentration was prescribed in their experiments, which therefore did not include the full interactions of aerosol radiation, microphysics, and removal processes, with monsoon dynamics. Jin et al. (2015) using WRF-CHEM and satellite observations also found that the influx of Middle East dust, mainly through radiative forcing, could redistribute and increase Indian summer monsoon regional rainfall, consistent with our results. However the relative low resolution model (~35 km) they used would be unable to resolve the aerosol-monsoon dynamic interaction over the complex topography of Himalayas regions. Moreover, they focused on the mean response of the entire Indian monsoon season (June–August), while we emphasize the short-term (<7 days) interactions during the onset phase of the monsoon. We also note that, even with the 9-km resolution used our experiments, the resolution is still relative coarse, in terms of realistic representation of aerosol-cloud microphysics. Cloud resolving scales (<2–3 km) is highly desirable for better representation the interaction of aerosol with the complex topography of the Himalaya foothills. However, for such high resolution simulation, only short-term integrations, with single or few ensemble simulations are feasible, because of the prohibitive computational demands. Furthermore, as noted in the discussion in Sect. 3.2, while NUWRF simulated the AOD spatial distribution reasonably well, the magnitude of the aerosol loading was significantly underestimated, compared to MODIS observation. Assuming that a heavier aerosol loading will produce stronger impacts on monsoon, it is plausible that the present results may have underestimated the actual impacts of aerosols in reality.
As a caveat, we note that our results are based on a singular monsoon year of 2008. More studies need to be carried out to ascertain the robustness of the results. Nonetheless, the fundamental physical processes revealed by this study, together with many recent studies of aerosol impacts on monsoon weather and climate, lend support to the paradigm that aerosol from natural sources is an essential components of the Asian monsoon climate, affecting variability of the monsoon on diverse spatio-temporal scales (Lau 2016).
Finally, an important message from our study, consistent with many contemporary studies on aerosol-monsoon interactions (See Li et al. 2016 for a comprehensive review), is that aerosol–monsoon interaction is important not only for climate change but also for short-term weather (3–7 days) forecasts, and very likely for medium to long-term forecasts (>7 days) as well as seasonal-to-interannual predictions of the Asian monsoon.
This work is partially supported by the NASA Precipitating Measuring Mission (PMM) Program, the NUWRF project under NASA Modeling and Analysis and Prediction Program (MAP), and the Department of Energy/ Pacific Northwest National Laboratory Grant 4313671 to ESSIC, University of Maryland.
- Bosilovich M et al (2015) MERRA-2: initial evaluation of the climate. In: Koster R (ed) Technical Report Series on Global Modeling and Data Assimilation. NASA/TM-2015-104606/Vol.43Google Scholar
- Chin M, Chu DA, Levy R, Remer L, Kaufman Y, Holben B, Eck T, Ginoux P, Gao Q (2004) Aerosol distribution in the Northern Hemisphere during ACE-Asia: results from global model, satellite observations, and Sun photometer measurements. J Geophys Res 109:D23S90. doi: 10.1029/2004GL02014 CrossRefGoogle Scholar
- Chin M, Diehl T, Tan Q, Prospero JM, Kahn RA, Remer LA, Yu H, Sayer AM, Bian H, Geogdzhayev IV, Holben BN, Howell SG, Huebert BJ, Hsu NC, Kim D, Kucsera TL, Levy RC, Mishchenko MI, Pan X, Quinn PK, Schuster GL, Streets DG, Strode SA, Torres O, Zhao X-P (2014) Multi-decadal variations of atmospheric aerosols from 1980 to 2009: a perspective from observations and a global model. Atmos Chem Phys 14:3657–3690CrossRefGoogle Scholar
- Das S, Mohanty UC, Tyagi A, Sikka DR, Joseph PV, Rathore LS, Habib A, Baidya SK, Snam K, Sarkar A (2014) The SAARC STORM: a coordinated filed experiment on severe thunderstorm observations and regional modeling over the South Asian region. Bull Am Meteorol Soc. doi: 10.1175/BAMS-D-12-00237.1 Google Scholar
- Henriksson SV, Pietikäinen J-P, Hyvärinen A-P, Räisänen P, Kupiainen K, Tonttila J, Hooda R, Lihavainen H, O’Donnell D, Backman L, Klimont Z, Laaksonen A (2014) Spatial distributions and seasonal cycles of aerosol climate effects in India seen in a global climate–aerosol model. Atmos Chem Phys 14:10177–10192. www.atmos-chem-phys.net/14/10177/2014/ doi: 10.5194/acp-14-10177-2014
- Lau KM, Ramanathan V, Wu G-X, Li Z, Tsay SC, Hsu C, Sikka R, Holben B, Lu D, Tartari G, Chin M, Koudelova P, Chen H, Ma Y, Huang J, Taniguchi K, Zhang R (2008) the joint aerosol-monsoon experiment: a new challenge in monsoon climate research. Bull Am Meteorol Soc 89:369–383. doi: 10.1175/BAMS-89-3-369 CrossRefGoogle Scholar
- Li Z, Lau WK-M, Ramanathan V, Wu G, Ding Y, Manoj MG, Liu J, Qian Y, Li J, Zhou T, Fan J, Rosenfeld D, Ming Y, Wang Y, Huang J, Wang B, Xu X, Lee S-S, Cribb M, Zhang F, Yang X, Takemura T, Wang K, Xia X, Yin Y, Zhang H, Guo J, Zhai PM, Sugimoto N, Babu SS, Brasseur GP (2016) Aerosol and monsoon climate interactions over Asia. Geophys Rev 54. doi: 10.1002/2015RG00050
- Peters-Lidard PR, Houser Y, Tian SV, Kumar J, Geiger S, Olden L, Lighty B, Doty P, Dirmeyer J, Adams K, Michell EF, Wood J Sheffield (2007) High-performance earth system modeling with NASA/GSFC’s land information system (LIS). Innov Syst Softw Eng 3:157–165. doi: 10.1007/s11334-007-0028-x CrossRefGoogle Scholar
- Peters-Lidard CD, Kemp EM, Matsui T, Santanello JA Jr, Kumar SV, Jacob JP, Clune T, Tao W-K, Chin M, Hou A, Case JL, Kim D, Kim K-M, Lau W, Liu Y, Shi J-J, Starr D, Tan Q, Tao Z, Zaitchik BF, Zavodsky B, Zhang SQ, Zupanski M (2015) Integrated modeling of aerosol, cloud, precipitation and land processes at satellite-resolved scales. Environ Model Softw 67:145–159. doi: 10.1016/j.envsoft.2015.01.007 CrossRefGoogle Scholar
- Shi JJ, Matsui T, Tao W-K, Peters-Lidard C, Chin M, Tan Q, Pickering K, Guy N, Lang S, Kemp E (2014) Implementation of an aerosol-cloud microphysics-radiation coupling into the NASA unified WRF: simulation results for the 6–7 August 2006 AMMA special observing period. Q J R Meteorol Soc 140:2158–2175. doi: 10.1002/qj.2286 CrossRefGoogle Scholar
- Tyagi A, Hatwar HR, Pai DS (2009) Monsoon 2008: a report. National Climate Centre, Indian Met. Dept., Monograph, Synoptic Meterorology No: 07/2009Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.