- Open Access
Using remotely sensed and climate data to predict the current and potential future geographic distribution of a bird at multiple scales: the case of Agelastes meleagrides, a western African forest endemic
© The Author(s) 2019
- Received: 17 December 2018
- Accepted: 27 May 2019
- Published: 10 June 2019
Understanding geographic distributions of species is a crucial step in spatial planning for biodiversity conservation, particularly as regards changes in response to global climate change. This information is especially important for species of global conservation concern that are susceptible to the effects of habitat loss and climate change. In this study, we used ecological niche modeling to assess the current and future geographic distributional potential of White-breasted Guineafowl (Agelastes meleagrides) (Vulnerable) across West Africa.
We used primary occurrence data obtained from the Global Biodiversity Information Facility and national parks in Liberia and Sierra Leone, and two independent environmental datasets (Moderate Resolution Imaging Spectroradiometer normalized difference vegetation index at 250 m spatial resolution, and Worldclim climate data at 2.5′ spatial resolution for two representative concentration pathway emissions scenarios and 27 general circulation models for 2050) to build ecological niche models in Maxent.
From the projections, White-breasted Guineafowl showed a broader potential distribution across the region compared to the current IUCN range estimate for the species. Suitable areas were concentrated in the Gola rainforests in northwestern Liberia and southeastern Sierra Leone, the Tai-Sapo corridor in southeastern Liberia and southwestern Côte d’Ivoire, and the Nimba Mountains in northern Liberia, southeastern Guinea, and northwestern Côte d’Ivoire. Future climate-driven projections anticipated minimal range shifts in response to climate change.
By combining remotely sensed data and climatic data, our results suggest that forest cover, rather than climate is the major driver of the species’ current distribution. Thus, conservation efforts should prioritize forest protection and mitigation of other anthropogenic threats (e.g. hunting pressure) affecting the species.
- Climate change
- Conservation planning
- Ecological niche modeling
- Species distribution
- Upper Guinea Forest
- White-breasted Guineafowl
West Africa is recognized as a global biodiversity hotspot, yet the ecology, distributions, and population numbers of most of its biota remain largely unknown (Mittermeier et al. 1998; Myers et al. 2000), which makes it challenging to optimize conservation of threatened species in the region. The region also ranks among the most deforested places on Earth, and is considered as highly susceptible to impacts of global climate change (Stattersfield et al. 1998; Myers et al. 2000; Malcolm et al. 2006; Janes and Hartley 2010; Díaz et al. 2018; IPBES 2018). As such, understanding likely relative impacts of these changes on geographic distributions of the region’s biota is important for conservation planning and decision-making.
A research approach used to address such questions is ecological niche modeling, which integrates known occurrences of species and environmental variables (e.g. temperature, precipitation) to characterize potential future geographic distributions of species in response to global climate change (Pearson and Dawson 2003; Peterson et al. 2011). Ecological niche models are useful tools in predicting or identifying priority areas for research and conservation of species and particularly for poorly known species (Pearson et al. 2007; Guisan et al. 2013).
In this study, we used ecological niche modeling to assess the distributional status of a Vulnerable Forest Endemic bird species, the White-breasted Guineafowl (Agelastes meleagrides) across West Africa. White-breasted Guineafowl belongs to the family Numidae and order Galliformes. Galliformes are thought to be highly susceptible to climate change owing to their relative inability to disperse over large distances and their consequently high site-fidelity (WPA-IUCN 2009). Habitat loss and climate change both have drastic effects on the distributions and populations of birds (Şekercioğlu et al. 2012; Stephens et al. 2016).
Here, we aimed to (1) estimate the current geographic distribution of White-breasted Guineafowl, (2) assess impacts of global climate change on the current and potential future geographic distribution of the species, and (3) identify priority areas for research and conservation of the species.
This study was conducted across the Upper Guinea lowland rainforest of West Africa, which is one of two major forest blocks in West and Central Africa, a global biodiversity hotspot, and a BirdLife International-designated Endemic Bird Area. Primary occurrence data for White-breasted Guineafowl were obtained from two sources: the Global Biodiversity Information Facility (GBIF 2017) and surveys in two national parks in Liberia and Sierra Leone (Sapo National Park in Liberia and Gola Rainforest National Park in Sierra Leone). The data obtained were error-checked, for example, for precision in GPS coordinates and wrong arrangement of longitude or latitude, and improved where necessary to meet appropriate standards for ecological niche modeling as explained below. To this end, for GBIF data, we split the dataset into georeferenced and non-georeferenced portions. Georeferenced data were checked for errors and data consistency as regards geographic coordinates (Chapman 2005). Where possible, errors were fixed; otherwise, erroneous data points were removed from the datasets. Non-georeferenced data that had detailed both datasets were rarefied spatially to remove duplicate points, using SDM Toolbox v2.2b in ArcGIS v10.4.1 (Brown 2014). We split the two occurrence datasets into two portions each: 70% for model calibration and 30% for evaluation.
Next, we acquired two independent environmental datasets, remotely sensed data and climatic data, for building ecological niche models. For climatic data, we acquired 15 climatic variables at 2.5’ (~ 5 km) spatial resolution, characterizing temperature and precipitation and their seasonality for present-day climatic conditions (Hijmans et al. 2005). To characterize future climate conditions, we used data for two IPCC representative concentration pathway (RCP 4.5, 8.5) emissions scenarios and 27 general circulation models (GCMs), all for one future time period (2050; Hijmans et al. 2005; Additional file 1: Table S1). The RCPs summarize low and high emissions scenarios as regards greenhouse gases, and the GCMs represent independent simulations of global climate processes and thus allow understanding of uncertainty as regards future climate patterns. For remotely sensed data, we obtained Moderate Resolution Imaging Spectroradiometer (MODIS 2017) normalized difference vegetation index (NDVI) data layers at 250 m spatial resolution from 2012 to 2016 for the region. The MODIS data consist of gridded vegetation index maps that depict temporal and spatial variations in the earth’s vegetation activity (Didan et al. 2015). These vegetation indices are derived on 16-day and monthly intervals and provide consistent, spatial and temporal time series comparisons of global vegetation conditions that can be used to monitor the Earth’s terrestrial photosynthetic vegetation activity in support of phenologic, change detection, and biophysical interpretations (Huete et al. 1999; Didan et al. 2015). Environmental datasets were processed as explained below.
For the MODIS dataset, in which vegetation index data from successive time periods can be highly intercorrelated, we applied three principal components analysis approaches: T-mode, S-mode, and Fourier to reduce colinerity or spatial and temporal autocorrelation (Martins et al. 2012; Zhang et al. 2012; Gocic and Trajkovic 2014; Eastman 2015). In T-mode, each image in the series is considered to be an independent variable in the analysis, components are images and the loadings are graphs (Zhang et al. 2012). In contrast, in S-mode, each pixel is considered as an independent sample in space and time; components are graphs and the loadings are images (Martins et al. 2012; Gocic and Trajkovic 2014). Fourier approaches use time series transposed into a set of Fourier components (amplitudes and phases) and apply principal components analysis (Manjunath Aradhya et al. 2010; Sobrino and Julien 2013). Following the analysis, we selected three subsets each of T-mode, S-mode, and Fourier variables for modeling. Following the data processing for both occurrence and environmental datasets, we left with 67 occurrence records for MODIS data and 31 occurrence records for climatic data.
Ecological niche modeling
We used the Maxent algorithm to estimate the ecological niche of our target species (Phillips et al. 2006). To calibrate models and choose optimal parameter values, for climatic variables (coarse-scale data), we assessed 29 combinations of Maxent feature classes (all possible combinations of linear, quadratic, product, threshold, and hinge features) as they interacted with 17 regularization multiplier values (0.1, 0.2, 0.3…1, 2, 4, 5, 6, 8, and 10). For MODIS variables, we assessed a simpler set of five combinations of Maxent feature classes (linear, quadratic, product, threshold, and hinge) as they interacted with nine regularization multiplier values (0.1, 0.3, 0.5, 0.7, 1, 3, 5, 7, and 10) to speed up the processing time, given the high spatial resolution of this dataset. For both fine-scale and coarse-scale analyses, we split the occurrence data into two portions: 70% for model calibration and 30% for evaluation. We evaluated candidate models and selected best models using the Akaike information criterion corrected (AICc) for small sample sizes to evaluate model complexity (Hurvich and Tsai 1989), performed significance tests using partial ROC (Peterson et al. 2008), and evaluated performance using a 5% training presence threshold to evaluate omission (Peterson et al. 2011).
Finally, to generate binary maps of the present and future time periods, we transferred the final (“best”) models to the present and future (climatic data only). We applied a 5% training omission threshold. To assess the vulnerability of White-breasted Guineafowl to global climate change, we combined medians of the present and 27 future projections for each RCP to create a visualization that shows the potential for range expansion, range retraction, and range stability (Campbell et al. 2015). All analyses were performed in ArcGIS version 10.5.1 (ESRI 2017) and the “kuenm” R Package (Cobos et al. 2019).
In all, 405 candidate models were calibrated for White-breasted Guineafowl using the MODIS dataset. All 405 candidate models performed significantly better than random expectations (p < 0.001); 123 models met our 5% omission rate threshold; and five models were within two units of the minimum AICc scores. Of the 123 models with low omission rates, 75 were candidate models built with T-mode variables (including the five models within two units of the minimum AICc), whereas only 7 and 41 models were for Fourier and S-mode respectively. Of the five candidate T-mode models within two units of the minimum AICc and low omission rates, only one had zero omission rate. This model was selected as the best model and projected across the full extent of our region of interest (Fig. 2; Additional file 1: Table S2). From the model projections, White-breasted Guineafowl showed a broader distribution across its range compared to IUCN’s highly fragmented estimated range for the species (Figs. 1 and 2). Suitable areas in model outputs were concentrated in the Gola rainforest complex in northwestern Liberia and southeastern Sierra Leone; the Tai-Sapo complex in southeastern Liberia and southwestern Cote d’Ivoire; and the Nimba Mountains in northern Liberia, southeastern Guinea, and northwestern Cote d’Ivoire (Fig. 2). Our projections also showed a significant extent of suitable habitat in southwestern Ghana and southeastern Cote d’Ivoire (Figs. 1, 2).
This study lays out the first detailed model-based projections of the geographic distribution of White-breasted Guineafowl across the Upper Guinea lowland rainforest region, providing a baseline for regional-to-local-level conservation planning and decision-making. Our results confirmed that GRNP and its sister reserve Gola Forest National Park (GFNP) in Liberia hold broad suitable areas for White-breasted Guineafowl, emphasizing the importance of these sites in the conservation of the westernmost range of the species. The good news here is that these two areas are designated protected areas in both countries. The decree of GFNP by the Liberian government in 2016 created a transboundary landscape or corridor between Liberia and Sierra Leone to increase the dispersal potential of species in general, including, the White-breasted Guineafowl within a protected environment across the Gola forest landscape.
Additionally, we acknowledge current efforts to create a network of protected areas across the region to increase transboundary landscape connectivity in Liberia, Sierra Leone, Guinea, and Cote d’Ivoire (e.g. GRNP Sierra Leone/GFNP Liberia; Ziama Guinea/Wonogizi Liberia; Tai Côte d’Ivoire/Sapo-Grebo Liberia) that will optimize overall biodiversity conservation across the region. However, for threatened species, such efforts need to be extended to fragmented isolated populations/habitats outside these targeted landscapes (Freeman et al. 2019a, b). For instance, the easternmost population of White-breasted Guineafowl in southwestern Ghana and southeastern Côte d’Ivoire is isolated from most of the species’ known populations in the west.
Unexpectedly the projected impacts of climate change on the geographic distribution of White-breasted Guineafowl were minimal, suggesting stability across the species’ range for the present and in the future, at least as regards climate change effects. Low sensitivity to climate change in this species does match the general observation for West African birds (Carr et al. 2014; Baker et al. 2015). However, although minimum, the notable projected coastal range loss of the species should be considered in any future conservation planning. This potential range loss could be attributed to increasing projected global sea level rise that would increase the chances of coastal erosion, thereby changing microclimatic conditions. Liberia is already experiencing rapid coastal erosion due to global sea level rise. Galliformes are thought to be at most risk from climate induced changes in higher elevations (Li et al. 2010); the low elevation habitat in West Africa inhabited by White-breasted Guienafowl will not face such extreme climate shifts. However, an important caveat is that the potential behavioral and demographic impacts of climate change on West African birds remain largely unexplored. For example, the cues for the timing of breeding for Black Grouse (Tetrao tetrix) in Finland have been shifted, such that the survival probability of chicks is much lower, as they are exposed to lower temperatures as the breeding cues shift further into the summer (Ludwig et al. 2006). These demographic consequences of climate change need to be investigated fully for the White-breasted Guineafowl and other West African taxa.
A prominent observation in our results, however, was the projected westwards potential climate-change-driven range expansion of the species, extending to the Western Peninsula National Park in southwestern Sierra Leone. We know of no record of the species from this site, but it already offers suitable forested habitat for some of the Upper Guinea Forest endemics (e.g. White-necked Picathartes Picathartes gymnocephalus). However, given the projected potential connectivity between Western Peninsula National Park and GRNP, under favorable climatic conditions in the future, this forest complex could exemplify a classic metapopulational dynamics, with GRNP providing viable source populations of White-breasted Guineafowl to colonize the site.
In conclusion, we show that White-breasted Guineafowl has a broader present-day distribution than represented by IUCN (2018), and that projected climate-change-driven impacts on the species are minimal. This latter result suggests that forest cover, and not necessarily climate change, is the most important driver of the species’ geographic distribution. As such, national and regional conservation efforts should prioritize forest protection and anthropogenic threats (e.g. hunting pressure, shifting cultivation) impacting the species. This new perspective on the species’ distribution is important for its conservation given increasing threats across its range. It provides an opportunity to include isolated populations that were originally overlooked for lack of information. These results demonstrate that by combining remotely sensed data and climatic data in ecological niche modeling, we understand not only species’ responses to global climate, but also their responses to land use and land cover.
We are thankful to A. Townsend Peterson for his helpful review and comments during the preparation of this manuscript and to members of the KUENM group at the University of Kansas for their useful feedback on methods used in this study.
This project was supported by Conservation International through a Global Environment Facility-funded Grant No. GEF-5810.
BF and MG conceived of the project ideas and planned the study. BB collected data. BF and DJG conducted research and data analysis. BF, MG, BB, and DJG composed manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
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. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Allport G. The status and conservation of threatened birds in the Upper Guinea forest. Bird Conserv Int. 1991;1:53–74.View ArticleGoogle Scholar
- Baker DJ, Hartley AJ, Burgess ND, Butchart SHM, Carr JA, Smith RJ, Belle E, Willis SG. Assessing climate change impacts for vertebrate fauna across the West African protected area network using regionally appropriate climate projections. Divers Distrib. 2015;21:991–1003.View ArticleGoogle Scholar
- Barve N, Barve V, Jiménez-Valverde A, Lira-Noriega A, Maher SP, Peterson AT, Soberón J, Villalobos F. The crucial role of the accessible area in ecological niche modeling and species distribution modeling. Ecol Model. 2011;222:1810–9.View ArticleGoogle Scholar
- Brown JL. SDMtoolbox: a python-based GIS toolkit for landscape genetic, biogeographic and species distribution model analyses. Methods Ecol Evol. 2014;5:694–700.View ArticleGoogle Scholar
- Campbell LP, Luther C, Moo-Llanes D, Ramsey JM, Danis-Lozano R, Peterson AT. Climate change influences on global distributions of dengue and chikungunya virus vectors. Philos Trans R Soc B. 2015;370:20140135.View ArticleGoogle Scholar
- Carr JA, Hughes AF, Foden WB. A climate change vulnerability assessment of West African species. Cambridge: Technical Report, UN Environment World Conservation Monitoring Centre; 2014.Google Scholar
- Chapman AD. Principles of data cleaning–primary species and species occurrene data, version 1.0. Copenhagen: Report for the Global Biodiversity Information Facility. http://www.gbif.org/resource/80528 (2005).
- Cobos ME, Peterson AT, Barve N, Osorio-Olvera L. kuenm: an R package for detailed development of ecological niche models using Maxent. PeerJ. 2019;7:e6281.View ArticleGoogle Scholar
- Davies AG. The Gola forest reserves, Sierra Leone. Wildlife conservation and forest management. Gland: International Union for Conservation of Nature and Natural Resources; 1987.Google Scholar
- Díaz S, Pascual U, Stenseke M, Martín-López B, Watson RT, Molnár Z, Hill R, Chan KMA, Baste IA, Brauman KA, et al. Assessing nature’s contributions to people. Science. 2018;359:270–2.View ArticleGoogle Scholar
- Didan K, Munoz AB, Solano R, Huete A. MODIS vegetation index user’s guide (MOD13series). Vegetation Index and Phenology Lab, The University of Arizona; 2015. p. 1–38.Google Scholar
- Dowsett-Lemaire F, Dowsett RJ. The avifauna of the proposed Kyabobo National Park in eastern Ghana. Malimbus. 2007;29:61–88.Google Scholar
- Eastman JR. Accessed in TerrSet version, 18.07. Worcester: Clark Labs, Clark University; 2015.Google Scholar
- ESRI. ArcGIS Release 10.5.1. Redlands: Environmental Systems Research Institute; 2017.Google Scholar
- Francis IS, Penford N, Gartshore ME, Jaramillo A. The White-breasted Guineafowl Agelastes meleagrides in Taï National Park, Côte d’Ivoire. Bird Conserv Int. 1992;2:25–60.View ArticleGoogle Scholar
- Freeman B, Roehrdanz PR, Peterson AT. Modeling endangered mammal species distributions and forest connectivity across the humid Upper Guinea lowland rainforest of West Africa. Biodivers Conserv. 2019a;28:671–85.View ArticleGoogle Scholar
- Freeman B, Danjuma DF, Molokwu-Odozi M. Status of globally threatened birds of Sapo National Park, Liberia. Ostrich. 2019b;90:19–24.View ArticleGoogle Scholar
- Fuller RA, Garson PJ. Pheasants: status survey and conservation action plan 2000‒2004. UK: IUCN; 2000.Google Scholar
- GBIF (Global Biodiversity Information Facility Database). GBIF occurrence download. https://doi.org/10.15468/dl.jdkowf (2017). Accessed 28 Nov 2017.
- Gocic M, Trajkovic S. Spatiotemporal characteristics of drought in Serbia. J Hydrol. 2014;510:110–23.View ArticleGoogle Scholar
- Guisan A, Tingley R, Baumgartner JB, Naujokaitis-Lewis I, Sutcliffe PR, Tulloch AIT, Regan TJ, Brotons L, McDonald-Madden E, Mantyka-Pringle C, et al. Predicting species distributions for conservation decisions. Ecol Lett. 2013;16:1424–35.View ArticleGoogle Scholar
- Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A. Very high resolution interpolated climate surfaces for global land areas. Int J Climatol. 2005;25:1965–78.View ArticleGoogle Scholar
- Huete A, Justice C, Van Leeuwen W. MODIS vegetation index (MOD13). Algorithm Theor Basis Doc. 1999;3:213.Google Scholar
- Hurvich CM, Tsai C-L. Regression and time series model selection in small samples. Biometrika. 1989;76:297–307.View ArticleGoogle Scholar
- IPBES. Assessment report on biodiversity and ecosystem services for Africa. Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. https://www.ipbes.net/assessment-reports/africa (2018).
- IUCN. The IUCN red list of threatened species. http://www.iucnredlist.org (2018).
- Janes T, Hartley A. Regional climate projections for West Africa. Cambridge: Technical Report, UN Environment World Conservation Monitoring Centre; 2010.Google Scholar
- Klop E, Lindsell JA, Siaka AM. The birds of Gola Forest and Tiwai Island, Sierra Leone. Malimbus. 2010;32:33–58.Google Scholar
- Li R, Tian H, Li X. Climate change induced range shifts of Galliformes in China. Integr Zool. 2010;5:154–63.View ArticleGoogle Scholar
- Ludwig GX, Alatalo RV, Helle P, Lindén H, Lindström J, Siitari H. Short-and long-term population dynamical consequences of asymmetric climate change in black grouse. Phil Trans R Soc B. 2006;273:2009–16.Google Scholar
- Malcolm JR, Liu C, Neilson RP, Hansen L, Hannah LEE. Global warming and extinctions of endemic species from biodiversity hotspots. Conserv Biol. 2006;20:538–48.View ArticleGoogle Scholar
- Manjunath Aradhya VN, Niranjan SK, Hemantha Kumar G. Probabilistic neural network based approach for handwritten character recognition. Int J Comput Commun. 2010;1:9–13.Google Scholar
- Martins DS, Raziei T, Paulo AA, Pereira LS. Spatial and temporal variability of precipitation and drought in Portugal. Nat Hazard Earth Syst. 2012;12:1493–501.View ArticleGoogle Scholar
- Mittermeier RA, Myers N, Thomsen JB, Da Fonseca GA, Olivieri S. Biodiversity hotspots and major tropical wilderness areas: approaches to setting conservation priorities. Conserv Biol. 1998;12:516–20.View ArticleGoogle Scholar
- MODIS Database. https://modis.gsfc.nasa.gov/data. Accessed 25 Sept 2017.
- Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J. Biodiversity hotspots for conservation priorities. Nature. 2000;403:853–8.View ArticleGoogle Scholar
- Olson DM, Dinerstein E. The Global 200: priority ecoregions for global conservation. Ann Mo Bot Gard. 2002;89:199–224.View ArticleGoogle Scholar
- Pearson RG, Dawson TP. Predicting the impacts of climate change on the distribution of species: are bioclimate envelope models useful? Glob Ecol Biogeogr. 2003;12:361–71.View ArticleGoogle Scholar
- Pearson RG, Raxworthy CJ, Nakamura M, Townsend Peterson A. Predicting species distributions from small numbers of occurrence records: a test case using cryptic geckos in Madagascar. J Biogeogr. 2007;34:102–17.View ArticleGoogle Scholar
- Peterson AT, Papeş M, Soberón J. Rethinking receiver operating characteristic analysis applications in ecological niche modeling. Ecol Model. 2008;213:63–72.View ArticleGoogle Scholar
- Peterson AT, Soberón J, Pearson RG, Anderson RP, Martínez-Meyer E, Nakamura M, Araújo MB. Ecological niches and geographic distributions. Princeton: Princeton University Press; 2011.View ArticleGoogle Scholar
- Phillips SJ, Dudík M, Schapire RE. A maximum entropy approach to species distribution modeling. In: Proceedings of the twenty-first international conference on Machine learning Association for Computing Machinery; 2006. p. 83.Google Scholar
- Şekercioğlu ÇH, Primack RB, Wormworth J. The effects of climate change on tropical birds. Biol Conserv. 2012;148:1–18.View ArticleGoogle Scholar
- Shcheglovitova M, Anderson RP. Estimating optimal complexity for ecological niche models: a jackknife approach for species with small sample sizes. Ecol Model. 2013;269:9–17.View ArticleGoogle Scholar
- Sobrino JA, Julien Y. Time series corrections and analyses in thermal remote sensing. Thermal infrared remote sensing. Dordrecht: Springer; 2013. p. 267–85.View ArticleGoogle Scholar
- Stattersfield A, Crosby M, Long A, Wege D. Endemic bird areas of the world: priorities for biodiversity conservation. Cambridge: BirdLife International; 1998.Google Scholar
- Stephens PA, Mason LR, Green RE, Gregory RD, Sauer JR, Alison J, Aunins A, Brotons L, Butchart SHM, Campedelli T, et al. Consistent response of bird populations to climate change on two continents. Science. 2016;352:84–7.View ArticleGoogle Scholar
- Waltert M, Seifert C, Radl G, Hoppe-Dominik B. Population size and habitat of the White-breasted Guineafowl Agelastes meleagrides in the Taï region, Côte d’Ivoire. Bird Conserv Int. 2010;20:74–83.View ArticleGoogle Scholar
- WPA-IUCN. Guidelines for the re-introduction of Galliformes for conservation purposes. Gland: World Pheasant Association and International Union for Conservation of Nature; 2009.View ArticleGoogle Scholar
- Zhang JP, Zhu T, Zhang QH, Li CC, Shu HL, Ying Y, Dai ZP, Wang X, Liu XY, Liang AM, Shen HX, Yi BQ. The impact of circulation patterns on regional transport pathways and air quality over Beijing and its surroundings. Atmos Chem Phys. 2012;12:5031–53.View ArticleGoogle Scholar