Climate change at the ecosystem scale: a 50-year record in New Hampshire

Observing the full range of climate change impacts at the local scale is difficult. Predicted rates of change are often small relative to interannual variability, and few locations have sufficiently comprehensive long-term records of environmental variables to enable researchers to observe the fine-scale patterns that may be important to understanding the influence of climate change on biological systems at the taxon, community, and ecosystem levels. We examined a 50-year meteorological and hydrological record from the Hubbard Brook Experimental Forest (HBEF) in New Hampshire, an intensively monitored Long-Term Ecological Research site. Of the examined climate metrics, trends in temperature were the most significant (ranging from 0.7 to 1.3 °C increase over 40–50 year records at 4 temperature stations), while analysis of precipitation and hydrologic data yielded mixed results. Regional records show generally similar trends over the same time period, though longer-term (70–102 year) trends are less dramatic. Taken together, the results from HBEF and the regional records indicate that the climate has warmed detectably over 50 years, with important consequences for hydrological processes. Understanding effects on ecosystems will require a diversity of metrics and concurrent ecological observations at a range of sites, as well as a recognition that ecosystems have existed in a directionally changing climate for decades, and are not necessarily in equilibrium with the current climate.     

Multidecadal oscillatory behaviour of rainfall extremes in Europe

Many studies have observed changes in the frequency and intensity of precipitation extremes and floods during the last decade(s). Natural variability by climate oscillations partly determines the observed evolution of precipitation extremes. Based on a technique for the identification and analysis of changes in extremes, this paper shows that precipitation extremes have oscillatory behaviour at multidecadal time scales. The analysis is based on a unique dataset of 108 years of 10-minute precipitation intensities at Uccle (Brussels), not affected by instrumental changes. We also checked the consistency of the findings with long precipitation records at 724 stations across Europe and the Middle East. The past 100 years show for northwestern Europe, both in winter and summer, larger and more precipitation extremes around the 1910s, 1950–1960s, and more recently during the 1990s–2000s. The oscillations for southwestern Europe are anti-correlated with these of northwestern Europe, thus with oscillation highs in the 1930–1940s and 1970s. The precipitation oscillation peaks are explained by persistence in atmospheric circulation patterns over the North Atlantic during periods of 10 to 15 years.     

The absence of memory in the climatic forcing of glaciers

Glaciers respond to both long-term, persistent climate changes as well as the year-to-year variability that is inherent to a constant climate. Distinguishing between these two causes of length change is important for identifying the true climatic cause of past glacier fluctuations. A key step in addressing this is to determine the relative importance of year-to-year variability in climate relative to more persistent climate fluctuations. We address this question for European climate using several long-term observational records: a century-long, Europe-wide atmospheric gridded dataset; longer-term instrumental measurements of summertime temperature where available (up to 250 years); and seasonal and annual records of glacier mass balance (between 30 and 50 years). After linear detrending of the datasets, we find that throughout Europe persistence in both melt-season temperature and annual accumulation is generally indistinguishable from zero. The main exception is in Southern Europe where a degree of interannual persistence can be identified in summertime temperatures. On the basis of this analysis, we conclude that year-to-year variability dominates the natural climate forcing of glacier fluctuations on timescales up to a few centuries.     

A stochastic model for the analysis of maximum daily temperature

Accurately predicting precipitation trends is vital in the economic development of a country. Ground observed data from the Nigeria Meteorological Agency (NIMET) was analyzed to study the long-term spatio-temporal trends of rainfall on annual and seasonal scales for 23 stations in Nigeria during a 40-year period spanning from 1974 to 2013. After testing the presence of autocorrelation, Mann–Kendall (modified Mann–Kendall) test was applied to non-autocorrelated (autocorrelated) series to detect the trends in rainfall data. Theil and Sen’s slope estimator test was used to find the magnitude of change over a time period. Pettitt’s test, Standard Normal Homogeneity Test, and Buishand’s test were further used to test the homogeneity of the rainfall series. The results show an increasing trend in annual rainfall; however, only nine stations have a significant increase during the period of study. On the seasonal time scale, a significant increasing trend was observed in the pre- and post-monsoon seasons, while only nine stations show a significant increasing trend in monsoon rainfall and a significant decreasing trend in the winter rainfall over the last 40 years. During the study period, 15.4 and 13.90 % increase were estimated for annual and monsoonal rainfall, respectively. Furthermore, seven stations exhibit changes in mean rainfall while majority of the stations considered (Eighteen stations) exhibit homogeneous trends in annual and seasonal rainfall over the country. The performance of the different tests used in this study was consistent at the verified significance level.     

Variations in cereal crop phenology in Spain over the last twenty-six years (1986–2012)

Over recent years, the Iberian Peninsula has witnessed an increase both in temperature and in rainfall intensity, especially in the Mediterranean climate area. Plant phenology is modulated by climate, and closely governed by water availability and air temperature. Over the period 1986–2012, the effects of climate change on phenology were analyzed in five crops at 26 sites growing in Spain (southern Europe): oats, wheat, rye, barley and maize. The phenophases studied were: sowing date, emergence, flag leaf sheath swollen, flowering, seed ripening and harvest. Trends in phenological response over time were detected using linear regression. Trends in air temperature and rainfall over the period prior to each phenophase were also charted. Correlations between phenological features, biogeographical area and weather trends were examined using a Generalized Lineal Mixed Model approach. A generalized advance in most winter-cereal phenophases was observed, mainly during the spring. Trend patterns differed between species and phenophases. The most noticeable advance in spring phenology was recorded for wheat and oats, the “Flag leaf sheath swollen” and “Flowering date” phenophases being brought forward by around 3 days/year and 1 day/year, respectively. Temperature changes during the period prior to phenophase onset were identified as the cause of these phenological trends. Climate changes are clearly prompting variations in cereal crop phenology; their consequences could be even more marked if climate change persists into the next century. Changes in phenology could in turn impact crop yield; fortunately, human intervention in crop systems is likely to minimize the negative impact.     

Reconstructing snowmelt in Idaho’s watershed using historic streamflow records

In recent decades, a warming climate likely has accelerated the timing of spring snowmelt in the western United States; however, records of the timing of snowmelt typically only extend to the 1980s. Stream gage data can lengthen records of the timing of snowmelt back to the early 1900s, enhancing understanding of past, current, and future climate change on snowmelt-dominated watersheds and associated ecosystems. We used snowpack telemetry data and historic streamflow records to test reconstructions of final snowmelt dates using Short Time Fourier Transform (STFT) wavelet analysis of hydrographs. STFT reconstructions tested against known final snowmelt dates over the last ~25 years indicate final snowmelt can be determined within ±4 days ~95% of the time and within ±7 days 100% of the time. Comparison of the STFT method with the center of timing method indicates that in addition to reconstructing actual snowmelt dates (as opposed to dates associated with the center of timing of streamflow), the STFT method may limit interpretation errors associated with changes in discharge not related to snowmelt. Reconstructions of final snowmelt dates in the Idaho, U.S. study area show intervals of early snowmelt (1920s–1930s), later and less variable snowmelt (1940s–1970s), and both variable and early snowmelt (~1985–2007). Early and variable snowmelt during the last ~20 years is associated with large wildfires.     

Modeling soil thermal and hydrological dynamics and changes of growing season in Alaskan terrestrial ecosystems

Abundant evidence indicates the growing season has been changed in the Alaskan terrestrial ecosystems in the last century as climate warms. Reasonable simulations of growing season length, onset, and ending are critical to a better understanding of carbon dynamics in these ecosystems. Recent ecosystem modeling studies have been slow to consider the interactive effects of soil thermal and hydrological dynamics on growing season changes in northern high latitudes. Here, we develop a coupled framework to model these dynamics and their effects on plant growing season at a daily time step. In this framework, we (1) incorporate a daily time step snow model into our existing hydrological and soil thermal models and (2) explicitly model the moisture effects on soil thermal conductivity and heat capacity and the effects of active layer depth and soil temperature on hydrological dynamics. The new framework is able to well simulate snow depth and soil temperature profiles for both boreal forest and tundra ecosystems at the site level. The framework is then applied to Alaskan boreal forest and tundra ecosystems for the period 1923–2099. Regional simulations show that (1) for the historical period, the growing season length, onset, and ending, estimated based on the mean soil temperature of the top 20 cm soils, and the annual cycle of snow dynamics, agree well with estimates based on satellite data and other approaches and (2) for the projected period, the plant growing season length shows an increasing trend in both tundra and boreal forest ecosystems. In response to the projected warming, by year 2099, (1) the snow-free days will be increased by 41.0 and 27.5 days, respectively, in boreal forest and tundra ecosystems and (2) the growing season lengths will be more than 28 and 13 days longer in boreal forest and tundra ecosystems, respectively, compared to 2010. Comparing two sets of simulations with and without considering feedbacks between soil thermal and hydrological dynamics, our analyses suggest coupling hydrological and soil thermal dynamics in Alaskan terrestrial ecosystems is important to model ecosystem dynamics, including growing season changes.     

Mid-Holocene tropical Pacific climate state,annual cycle,and ENSO in PMIP2 and PMIP3

Using the Paleoclimate Modeling Inter-comparison Project Phase 2 and 3 (PMIP2 and PMIP3), we investigated the tropical Pacific climate state, annual cycle, and El Niño-Southern Oscillation (ENSO) during the mid-Holocene period (6,000 years before present; 6 ka run). When the 6 ka run was compared to the control run (0 ka run), the reduced sea surface temperature (SST) and the reduced precipitation due to the basin-wide cooling, and the intensified cross-equatorial surface winds due to the hemispheric discrepancy of the surface cooling over the tropical Pacific were commonly observed in both the PMIP2 and PMIP3, but changes were more dominant in the PMIP3. The annual cycle of SST was weaker over the equatorial eastern Pacific, because of the orbital forcing change and the deepening mixed layer, while it was stronger over the equatorial western pacific in both the PMIP2 and PMIP3. The stronger annual cycle of the equatorial western Pacific SST was accompanied by the intensified annual cycle of the zonal surface wind, which dominated in the PMIP3 in particular. The ENSO activity in the 6 ka run was significantly suppressed in the PMIP2, but marginally reduced in the PMIP3. In general, the weakened air-sea coupling associated with basin-wide cooling, reduced precipitation, and a hemispheric contrast in the climate state led to the suppression of ENSO activity, and the weakening of the annual cycle over the tropical eastern Pacific might lead to the intensification of ENSO through the frequency entrainment. Therefore, the two opposite effects are slightly compensated for by each other, which results in a small reduction in the ENSO activity during the 6 ka in the PMIP3. On the whole, in PMIP2/PMIP3, the variability of canonical (or conventional) El Niño tends to be reduced during 6 ka, while that of CP/Modoki El Niño tends to be intensified.     

Impacts of climate change on August stream discharge in the Central-Rocky Mountains

In the snowmelt dominated hydrology of arid western US landscapes, late summer low streamflow is the most vulnerable period for aquatic ecosystem habitats and trout populations. This study analyzes mean August discharge at 153 streams throughout the Central Rocky Mountains of North America (CRMs) for changes in discharge from 1950–2008. The purpose of this study was to determine if: (1) Mean August stream discharge values have decreased over the last half-century; (2) Low discharge values are occurring more frequently; (3) Climatic variables are influencing August discharge trends. Here we use a strict selection process to characterize gauging stations based on amount of anthropogenic impact in order to identify heavily impacted rivers and understand the relationship between climatic variables and discharge trends. Using historic United States Geologic Survey discharge data, we analyzed data for trends of 40–59 years. Combining of these records along with aerial photos and water rights records we selected gauging stations based on the length and continuity of discharge records and categorized each based on the amount of diversion. Variables that could potentially influence discharge such as change in vegetation and Pacific Decadal Oscillation (PDO) were examined, but we found that that both did not significantly influence August discharge patterns. Our analyses indicate that non-regulated watersheds are experiencing substantial declines in stream discharge and we have found that 89% of all non-regulated stations exhibit a declining slope. Additionally our results here indicate a significant (α?≤?0.10) decline in discharge from 1951–2008 for the CRMs. Correlations results at our pristine sites show a negative relationship between air temperatures and discharge and these results coupled with increasing air temperature trends pose serious concern for aquatic ecosystems in CRMs.     

Trend analysis of precipitation in Jharkhand State,India

Jharkhand is one of the eastern states of India which has an agriculture-based economy. Uncertain and erratic distribution of precipitation as well as a lack of state water resources planning is the major limitation to crop growth in the region. In this study, the spatial and temporal variability in precipitation in the state was examined using a monthly precipitation time series of 111 years (1901–2011) from 18 meteorological stations. Autocorrelation and Mann–Kendall/modified Mann–Kendall tests were utilized to detect possible trends, and the Theil and Sen slope estimator test was used to determine the magnitude of change over the entire time series. The most probable change year (change point) was detected using the Pettitt–Mann–Whitney test, and the entire time series was sub-divided into two parts: before and after the change point. Arc-Map 9.3 software was utilized to assess the spatial patterns of the trends over the entire state. Annual precipitation exhibited a decreasing trend in 5 out of 18 stations during the whole period. For annual, monsoon and winter periods of precipitation, the slope test indicated a decreasing trend for all stations during 1901–2011. The highest variability was observed in post-monsoon precipitation (77.87 %) and the lowest variability was observed in the annual series (15.76 %) over the 111 years. An increasing trend in precipitation in the state was found during the period 1901–1949, which was reversed during the subsequent period (1950–2011).     

The importance of record length in estimating the magnitude of climatic changes: an example using 175 years of lake ice-out dates in New England

Many studies have shown that lake ice-out (break-up) dates in the Northern Hemisphere are useful indicators of late winter/early spring climate change. Trends in lake ice-out dates in New England, USA, were analyzed for 25, 50, 75, 100, 125, 150, and 175 year periods ending in 2008. More than 100 years of ice-out data were available for 19 of the 28 lakes in this study. The magnitude of trends over time depends on the length of the period considered. For the recent 25-year period, there was a mix of earlier and later ice-out dates. Lake ice-outs during the last 50 years became earlier by 1.8 days/decade (median change for all lakes with adequate data). This is a much higher rate than for longer historical periods; ice-outs became earlier by 0.6 days/decade during the last 75 years, 0.4 days/decade during the last 100 years, and 0.6 days/decade during the last 125 years. The significance of trends was assessed under the assumption of serial independence of historical ice-out dates and under the assumption of short and long term persistence. Hypolimnion dissolved oxygen (DO) levels are an important factor in lake eutrophication and coldwater fish survival. Based on historical data available at three lakes, 32 to 46 % of the interannual variability of late summer hypolimnion DO levels was related to ice-out dates; earlier ice-outs were associated with lower DO levels.     

Interannual to decadal summer drought variability over Europe and its relationship to global sea surface temperature

Interannual to decadal variability of European summer drought and its relationship with global sea surface temperature (SST) is investigated using the newly developed self calibrated Palmer drought severity index (scPDSI) and global sea surface temperature (SST) field for the period 1901–2002. A European drought severity index defined as the average of scPDSI over entire Europe shows quasiperiodic variations in the 2.5–5 year band as well as at 12–13 years suggesting a possible potential predictability of averaged drought conditions over Europe. A Canonical Correlation Analysis between summer scPDSI anomalies over Europe and global SST anomalies reveals the existence of three modes of coupled summer drought scPDSI patterns and winter global SST anomalies. The first scPDSI-SST coupled mode represents the long-term trends in the data which manifest in SST as warming over all oceans. The associated long-term trend in scPDSI suggests increasing drought conditions over the central part of Europe. The second mode is related to the inter-annual ENSO and decadal PDO influence on the European climate and the third one captures mainly the drought pattern associated to Atlantic Multidecadal Oscillation. The lag relationships between winter SST and summer drought conditions established in this study can provide a valuable skill for the prediction of drought conditions over Europe on interannual to decadal time scales.     

Creation of a homogenous climate record for Widdybank Fell in the Upper Teesdale National Nature Reserve, northern England: 1968-2006

A homogenous climate record (1968-2006) is created for Widdybank Fell (515 m) in the Upper Teesdale National Nature Reserve in northern England, one of the longest high-elevation records in the U.K. Separate time series from Widdybank Fell (1968-1995) and nearby Hunt Hall (1996-2006) are combined using a single mobile automatic weather station (AWS) to calibrate between locations based on 5 years of measurements (2000-2005). After instrumental differences are eradicated, transfer functions are developed based on monthly temperature differences, median monthly ratios of wind speed and mean monthly precipitation totals. The resultant monthly time series show limited trends, although minimum temperatures have increased by 0.38°C/decade. There are no secular trends in mean wind speed or monthly precipitation anomalies, in agreement with other studies which suggest northern England is in a transition area between predicted wetting in northern Europe and drying further south.     

Climatic controls on Holocene lake-level changes in Europe

Sensitivity experiments with a simple water-balance model were used to constrain the possible climatic causes of distinct Holocene patterns of lake-level variation in different regions of Europe. Lakes in S Sweden were low at 9 ka, high around 6.5 ka, low again around 4 ka and are high now. Lakes in Estonia show similar but weaker trends. Lakes in S France were highest around 9 ka, lowest around 4 ka, intermediate now. Lakes in Greece were also high around 9 ka but continued rising until 7.5 ka, then fell gradually from 5 ka with a brief high phase around 3 ka, and are low now. The model was forced with insolation anomalies deduced from orbital variations, temperature anomalies inferred from the pollen record and cloudiness anomalies derived from changes in the position of the subtropical anticyclone (inferred from reconstructed changes in the equator-to-pole temperature gradient), in order to evaluate the effects of resultant evaporation changes on catchment water balance. The resulting simulated changes in runoff (precipitation minus actual evapotranspiration) were slight, and frequently opposite to the observed trends. Larger changes in precipitation are plausible and are required to explain the data. The required precipitation increase in N Europe from 9 ka (low) to 6 ka (high) is suggested by GCM experiments to have been a consequence of interacting insolation and residual ice-sheet effects on the atmospheric circulation over the North Atlantic. The explanation of other observed changes, including the drying trend during the Holocene in S Europe, has not been provided by GCM experiments to date. Explanations may lie in changes in mesoscale circulation, sea-surface temperature patterns and the coupling between these phenomena that may not follow orbital changes in any simple way. Correspondence to: SP Harrison     

A century of climate and ecosystem change in Western Montana: what do temperature trends portend?

The physical science linking human-induced increases in greenhouse gasses to the warming of the global climate system is well established, but the implications of this warming for ecosystem processes and services at regional scales is still poorly understood. Thus, the objectives of this work were to: (1) describe rates of change in temperature averages and extremes for western Montana, a region containing sensitive resources and ecosystems, (2) investigate associations between Montana temperature change to hemispheric and global temperature change, (3) provide climate analysis tools for land and resource managers responsible for researching and maintaining renewable resources, habitat, and threatened/endangered species and (4) integrate our findings into a more general assessment of climate impacts on ecosystem processes and services over the past century. Over 100 years of daily and monthly temperature data collected in western Montana, USA are analyzed for long-term changes in seasonal averages and daily extremes. In particular, variability and trends in temperature above or below ecologically and socially meaningful thresholds within this region (e.g., ?17.8°C (0°F), 0°C (32°F), and 32.2°C (90°F)) are assessed. The daily temperature time series reveal extremely cold days (≤??17.8°C) terminate on average 20 days earlier and decline in number, whereas extremely hot days (≥32°C) show a three-fold increase in number and a 24-day increase in seasonal window during which they occur. Results show that regionally important thresholds have been exceeded, the most recent of which include the timing and number of the 0°C freeze/thaw temperatures during spring and fall. Finally, we close with a discussion on the implications for Montana’s ecosystems. Special attention is given to critical processes that respond non-linearly as temperatures exceed critical thresholds, and have positive feedbacks that amplify the changes.     

A comparison among olive flowering trends in different Mediterranean areas (south-central Italy) in relation to meteorological variations

Phenological observations of the anthesic phases of olive flowering in a central Mediterranean area were recorded over a 9-year period. The aim of this research was to compare the flowering dates in relationship to the meteorological changes. Pollen emission from anthers was monitored by remote instrumentation placed directly in olive groves and phenological data regarding daily pollen concentrations (pollen/m³) were recorded using a pollen monitoring methodology. The rhythm of the phenological phases emerged as dependent on the meteorological trend of the spring forcing temperature. Generally, the phenomenon of pollen emission occurred progressively earlier prior to 2001, while in the following 5 years, the trend seemed to be inverted, showing a progressive delay of flowering. The spring quarterly mean temperature trends registered by GISS data in Europe confirmed the presence of diverse meteorological behavior during the study period, probably causing the biological divergences that were monitored. The principal result of the present contribution is to evidence the relativity of empirical investigations and observations considering different time intervals. This is due to the partial, brief series (9 years) of flowering dates which have to be considered as part of a longer series (26 years) in order to have a complete vision of the true phenomenon.     

Spatiotemporal variations of potential evapotranspiration and aridity index in relation to influencing factors over Southwest China during 1960–2013

This study investigated the spatial–temporal patterns and trends of potential evapotranspiration (ET₀) and aridity index (AI) over Southwest China during 1960–2013 based on daily temperature, precipitation, wind speed, sunshine duration, total solar radiation, and relative humidity data from 108 meteorological stations. The Penman–Monteith model, Mann–Kendall (M–K) test, moving t test, and Morlet wavelet method were used. The results indicated that ET₀ and AI across the region displayed decreasing trends, but the former was significant. After 2000, regionally average trends in ET₀ and AI increased rapidly, indicating that droughts increased over Southwest China in recent years. Spatially, the changes of ET₀ and AI were dissimilar and not clustered, either. Temporally, both ET₀ and AI displayed obvious abrupt change points over different timescales and that of AI was during the winter monsoon period. Significant periodic variations with periods of 27, 13, and 5 years were found in ET₀, but only of 13 and 5 years existed in AI. Correlation analysis revealed that the sunshine duration and wind speed were the dominant factors affecting ET₀ and that AI showed strong negative correlation with precipitation. The findings of this study enhance the understanding of the relationship between climate change and drought in Southwest China, while the mechanism controlling the variation in drought requires further study.     

Temperature trends of Chennai City, India

Chennai is the fourth largest metropolitan city in India, and it is one of India's chief industrial and economic growth centers. The temperature change in Chennai is studied in this research by analyzing the mean maximum temperature (MMaxT), mean minimum temperature (MMinT), and mean annual temperature (MAT) from 1951 to 2010. Data are analyzed in three parts by running linear regression and by taking anomalies of all time periods: (a) the whole period from 1951 to 2010; (b) phase 1, 1951–1980; and (c) phase 2, 1981–2010. The trends have been evaluated by Student's t statistics and supported by Mann Kendall rank statistics. The observed change in temperature is positive, which has been clear increasing trends in MMaxT, MMinT, and MAT. MAT has increased 1.3°C since the last 60 years. MMaxT has increased up to 1.6°C, in which the second phase accounts for 75 % of the total change during the last 60 years. MMinT over Chennai has increased 1.0°C. There is a high rise in temperature during winter season.     

Temporal and spatial variability of rainfall over Greece

Recent studies have showed that there is a significant decrease in rainfall over Greece during the last half of the pervious century, following an overall decrease of the precipitation at the eastern Mediterranean. However, during the last decade an increase in rainfall was observed in most regions of the country, contrary to the general circulation climate models forecasts. An updated high-resolution dataset of monthly sums and annual daily maxima records derived from 136 stations during the period 1940–2012 allowed us to present some new evidence for the observed change and its statistical significance. The statistical framework used to determine the significance of the slopes in annual rain was not limited to the time independency assumption (Mann-Kendall test), but we also investigated the effect of short- and long-term persistence through Monte Carlo simulation. Our findings show that (a) change occurs in different scales; most regions show a decline since 1950, an increase since 1980 and remain stable during the last 15 years; (b) the significance of the observed decline is highly dependent to the statistical assumptions used; there are indications that the Mann-Kendall test may be the least suitable method; and (c) change in time is strongly linked with the change in space; for scales below 40 years, relatively close regions may develop even opposite trends, while in larger scales change is more uniform.     

The effect of sea surface temperature bias in the PMIP2 AOGCMs on mid-Holocene Asian monsoon enhancement

The effect of bias on control simulation is a significant issue for climate change modeling studies. We investigated the effect of the sea surface temperature (SST) bias in present day (0 ka) Atmosphere–Ocean Coupled General Circulation Model (AOGCM) simulations on simulations of the mid-Holocene (6 ka, i.e., 6,000 years before present) Asian monsoon enhancement. Because changes in ocean heat transport have a negligible effect on the 6 ka Asian monsoon (Ohgaito and Abe-Ouchi in Clim Dyn 29(1):39–50, 2007), we used an Atmospheric General Circulation Model (AGCM) rather than an AOGCM. Simulations using the AGCM coupled to a mixed layer ocean model (MLM) were conducted for 0 and for 6 ka with different ocean heat transport estimated from the climatological SST of the 0 ka simulations from nine Paleoclimate Modeling Intercomparison Project (PMIP) phase 2 (PMIP2) AOGCMs (henceforth “MA” is used to refer to experiments using the AGCM coupled with the MLM). No correlation between MA and the corresponding PMIP2 was seen in the 0 ka precipitation and it was not very strong for the 6 ka precipitation enhancement. Thus, the influences from the different AGCMs play a substantial role on the 0 ka precipitation and the 6 ka precipitation enhancement. The sensitivity experiments indicated that it was the pattern of the 0 ka SST bias which played a dominant role in the 0 ka precipitation and the 6 ka precipitation enhancement, not the difference in the mean value of the SST bias. The distributions of the 6 ka precipitation enhancements for the nine PMIP2 AOGCMs and nine MA experiments were compared. These showed that the effects of SST bias on 6 ka precipitation enhancement among the AOGCMs were not negligible. The effects of biases among the AGCMs were not negligible either, but of comparable size. That is, improvements in both the SST bias and the AGCM contribute to simulate better 6 ka monsoon.