Spatial Heating Monthly Degree-Day Features and ClimatologicPatterns in Turkey

Summary  Degree-days as a measure of accumulated temperature deviations from a base temperature have many practical applications in various human related activities such as home cooling, heating, plant growth in agriculture and power generation in addition to energy requirement. Long temperature records are necessary for their reliable estimations at given stations. In this paper, degree-day measure has been applied to monthly temperature records for systematically changed base temperature values from − 25 °C to + 35 °C with 5 °C increments at 255 meteorology stations in Turkey. The results are represented in the form of spatial degree-day distribution maps, which are then related to various climatic, meteorological and topographic features of Turkey. For instance, free surface water bodies in forms of surrounding seas, lakes and rivers insert retardation in the expansion of heating degree-days over large regions. On the other hand, cold air penetration from polar regions in the northeastern Turkey originating from Siberia appears at moderate base temperature heating degree-days. Received August 20, 1998 Revised June 21, 1999     

The Characteristics of Cold Air Outbreaks over the Eastern Highlands of Kenya

Summary Climatological statistics of extreme temperature events over Kenya are established from the analysis of daily and monthly maximum temperatures for a representative station (Nairobi Dagoretti Corner) over the period 1956–1997. The months of June to August were shown to be the coldest with a mean monthly maximum temperature of less than 22 °C. Seasonal (June to August) mean maximum temperature was 21.5 °C. Using this seasonal mean temperature for the period 1967–1997 delineated 1968 as the coldest year in this series and 1983 as the warmest year. Spectral analysis of the seasonal data, for both the coldest and the warmest years, revealed that the major periods were the quasi-biweekly (10 days) and the Intraseasonal Oscillations (23 days). Secondary peaks occurred at periods of 4–6 and 2.5–3.5 days. A temperature threshold of 16.7 °C during July was used to define cold air outbreaks over Nairobi. This threshold temperature of 16.7 °C was obtained from the mean July maximum temperature (20.9 °C) minus two standard deviations. Notable trends include a decrease in the frequency of station-days, between 1956 and 1997, with temperatures less than 16.7 °C during July. Surface pressure patterns indicate that the origin of the cold air is near latitude 25° S and to the east of mainland South Africa. The cold air near 25° S is advected northwards ahead of the surface pressure ridge. Received July 19, 1999 Revised January 11, 2000     

North Sea – Caspian Pattern (NCP) – an upper level atmospheric teleconnection affecting the eastern Mediterranean – implications on the regional climate

Summary ?A calendar of the negative and positive phases of the North Sea – Caspian Pattern (NCP) for the period 1958–1998 was used to analyse the implication of the NCP upper level teleconnections on the regional climate of the eastern Mediterranean basin. Series of monthly mean air temperature and monthly total rainfall from 33 stations across Greece, Turkey and Israel, for the same period, were used. For each month, from October to April, averages of the monthly mean temperatures and the monthly rainfall totals as well as the standardized values of both parameters were calculated separately for the negative (NCP (−)) and the positive (NCP (+)) phases of the NCP. At all stations and in all months, temperature values were significantly higher during the NCP (−) as compared with the NCP (+). Furthermore, apart from very few exceptions, the absolute monthly mean maximum and monthly mean minimum values were obtained during the NCP (−) and the NCP (+) phases, respectively. The maximum impact of the NCP on mean air temperature was detected in the continental Anatolian Plateau, where the mean seasonal differences are around 3.5 °C. This influence decreases westwards and southwards. The influence on the rainfall regime is more complex. Regions exposed to the southern maritime trajectories, in Greece and in Turkey, receive more rainfall during the NCP (−) phase, whereas in the regions exposed to the northern maritime trajectories, such as Crete in Greece, the Black Sea region in Turkey, and in all regions of Israel, there is more rainfall during the NCP (+) phase. The accumulated rainfall differences between the two phases are over 50% of the seasonal average for some stations. A comparison of the capabilities of the NCP, the North Atlantic Oscillation (NAO) and the Southern Oscillation (SO) indices to differentiate between below and above normal temperatures was made. The results have placed the NCP, as the best by far of all three teleconnections in its ability to differentiate between below or above normal temperatures and as the main teleconnection affecting the climate of the Balkans, the Anatolian Peninsula and the Middle East. These results may serve to downscale General Circulation Model (GCM) scenarios to a regional scale and provide forecasts regarding eventual temperature and/or precipitation changes. Received June 25, 2001; revised February 25, 2002; accepted March 3, 2002     

Influence of vegetation changes during the Last Glacial Maximum using the BMRC atmospheric general circulation model

 The influence of different vegetation distributions on the atmospheric circulation during the Last Glacial Maximum (LGM, 21 000 years before present) is investigated. The atmospheric general circulation model of the Bureau of Meteorology Research Center was run using a modern vegetation and in a second experiment with a vegetation reconstruction for the LGM. It is found that a change from conifer to desert and tundra causes an additional LGM cooling of 1–2 °C in Western Europe, up to −4 °C in North America and −6 °C in Siberia. An expansion of dryland vegetation causes an additional annual cooling of 1–2 °C for Australia and northern Africa. On the other hand, an increase of temperature (2 °C) is found in Alaska due to changes in circulation. In the equatorial region the LGM vegetation leads to an increased modelled temperature of 0.5–1.5 °C and decreased precipitation (30%) over land due to a reduction of the tropical rainforest, mainly in Indonesia, where the reduction of precipitation over land is associated with an increase of precipitation of 30% over the western Pacific. Received: 15 December 1999 / Accepted: 10 January 2001     

Variations of 500 hPa flow patterns over Iranand surrounding areas and their relationship with the climate of Iran

Summary In order to explore the spatial and temporal variations of 500 hPa flow patterns and their relationship with the climate of Iran, monthly mean geopotential heights for the region 0° E to 70° E and 20° N to 50° N, at 5 degree resolution, were analysed. The study period covered the winter months October to March during the period 1961–90. The monthly height of the 500 hPa level was averaged along each meridian from 25° N to 45° N. The height of the mean monthly pressure pattern was mapped against the study years. The results showed that the characteristics of the 500 hPa flow pattern varied over monthly and annual time scales. Principal Component Analysis, with S-mode and Varimax rotation, was also used to reduce the gridded data to 5 (6 in October) significant factors. The factor scores for each month were then correlated with monthly Z-scores of precipitation and temperature anomalies over Iran. The results showed that troughs and ridges located close to Iran had more influence on the climate of Iran. Two troughs were identified and named the Caspian and Syrian troughs. Received April 12, 2001 Revised July 24, 2001     

Climatic impacts of historical wetland drainage in Switzerland

The effects of historical land-use and land-cover changes on the climate of the Swiss Plateau in the different seasons were investigated. In the 19th century, a civil engineering project was initiated to reshape the lake and river system on the Swiss Plateau in order to ban the frequent flooding during extreme weather events. The landscape modifications consisted primarily of a conversion of wetlands with extended peat soils into a highly productive agricultural landscape. Historical maps (1800–1850) served as a basis for the reconstruction of the past land use. The “Lokal-Modell” of the Consortium for Small-Scale Modelling was used to conduct eight one-month long high-resolution simulations (1.5 × 1.5 km²) with present and past landscape conditions. The modified soil and surface properties led to distinctly altered energy and moisture exchanges at the surface and as a consequence affected the local and regional climate. The climatic changes show different characteristics and magnitudes in the cold (October – March) as compared to the warm season (April – September). The landscape modifications led to an average daytime cooling between −0.12 °C (January) and −0.61 °C (April) and a night-time warming of 0.19 °C−0.34 °C. The differences in the mean monthly temperatures show a warming of 0.1 °C−0.2 °C in the cold season and a cooling of similar magnitude in most of the study area in the warm season. The modification of the radiation budget and the surface energy balance distinctly affected the convective activity in the study area in the warm season, but had only a weak effect on convectivity in the cold season. The cloud coverage in the warm season is therefore distinctly reduced compared to the past.     

The relationship between lightning activity and surface wet bulb temperature and its variation with latitude in Australia

Summary Convective Available Potential Energy (CAPE) is the driving force for thunderstorm development. CAPE is closely controlled by wet bulb temperature. The lightning activity measured by a network of ten lightning flash counters widely distributed across continental Australia was studied as a function of wet bulb temperature. At each of the stations, the monthly total of lightning ground flashes, N, increased sharply with the increase of the monthly mean daily maximum wet bulb temperature, T_w, max. The dependence was strongest in the tropics and became less pronounced at temperate latitudes. In Darwin (latitude 12° S), the lightning ground flash activity increased by over three orders of magnitude over a 7 °C range of T_w, max. The corresponding increases for Coffs Harbour (latitude 30° S) and for Melbourne (latitude 38° S) were about one and a half orders of magnitude and about half an order of magnitude, respectively, each over a 10 °C range of T_w, max. Power law approximations were derived for each of the ten stations and showed that the logarithm of N was directly proportional to the power, P, of T_w, max. The value of P showed a sharp exponential decrease with increasing latitude away from the equator.     

Holocene climatic change in Morocco: a quantitative reconstruction from pollen data

 Annual precipitation, July and January temperatures were reconstructed from a continuous Holocene pollen sequence from the Middle Atlas, Morocco, using the best modern analogues method. The reconstructions show a clear difference between the early and late Holocene: from ∼10 ka to ∼6.5 ka the climate was drier and warmer than during the period since 6.5 ka. The average value of annual precipitation was ∼870 mm until 6.5 ka, then rose to ∼940 mm. Between 10 ka and 6.5 ka January and July temperatures were about 4 °C higher than the present. Both temperatures show a marked decrease between 7 ka and 6 ka. After 6.5 ka July and January temperatures fluctuated between 21 and 23 °C, and 2.5 and 5 °C respectively. January temperatures show a period of intermediate values (∼3.5 °C) between 4 ka and 5.5 ka. The reconstructed climate values generally match palaeolimnological data from the same core, which show five intervals of low lake level during the Holocene. They are also consistent with regional-scale COHMAP simulated palaeoclimate that shows contrasting patterns of rainfall variation between the northwesternmost part of Africa and the intertropical band. Received: 7 July 1997 / Accepted: 28 May 1998     

Sea level pressure patterns associated with dry or wet monthly rainfall conditions in Turkey

Summary  Monthly rainfall totals at 7 stations across Turkey and sea level pressure (SLP) in 16 grid points in the region delimited by the 20° E and 50° E longitudes and by the 30° N and 45° N latitudes were analysed. Data were available for a period longer than sixty years. The standard deviations of SLP at each grid point for each month, were calculated and mapped. For each station, months were defined as dry or wet according to their z scores: ≤ −1.0 or ≥ 1.0 respectively. Maps showing the SLP z scores of the corresponding dry or wet months for each station were prepared. The maps, enable to distinguish between SLP patterns associated with dry or wet conditions. Furthermore, correlations between monthly rainfall in each of the stations and SLP at each grid point were performed. The correlation coefficients were mapped. (a) The variability of the SLP decreases from the Balkans towards the Arabian Peninsula and is much larger in winter as compared with summer. (b) Relationship between rainfall in Turkey and the regional SLP is large in winter and non existing in summer. (c) Pressure patterns associated with dry conditions, show usually positive SLP departures, whereas, pressure patterns associated with wet conditions show usually negative SLP departures. (d) There is a great resemblance between pressure patterns associated with wet conditions and correlation maps of the same months. Received September 4, 2000 Revised January 15, 2001     

Variations in the Temperature Regime Across the Mediterranean During the Last Century and their Relationship with Circulation Indices

Summary  Circulation types were identified by means of zonal and meridional indices calculated separately over ten different regions of 20°×20° over the Mediterranean and Europe. Seasonal temperature trends in 22 grid boxes of 5°×5° covering the entire Mediterranean, and at six stations Lisbon, Madrid, Florence, Luqa (Malta), Athens and Jerusalem, were calculated. A warming trend in the period 1873–1989 was detected. The warming is more evident in the western Mediterranean with an average rate of about 0.4 [°C/100 yr], than in the eastern Mediterranean with an increase of only 0.2 [°C/100 yr]. A cooling trend in autumn in the eastern Mediterranean with an average rate of −0.5 [°C/100 yr] was detected and attributed to an increase in northerly meridional circulation in that region. Warming trends at Lisbon, Madrid, Florence, Athens and Jerusalem, were more important than the trends in the grid boxes containing these stations. This rapid warming was attributed to urban effects. No such effects were found in Luqa due to its location and the lack of urban effects there. Temperatures at Luqa, Athens and Jerusalem are highly positively correlated. Likewise, temperatures at Lisbon and Madrid. Temperatures at Florence are either correlated with Madrid or with Luqa. Negative or no correlations were found between Lisbon or Madrid with Athens or Jerusalem, except during the winter. This was attributed to the fact that favourable circulation for high temperatures in the eastern stations was opposite to the favourable circulation for high temperatures in the western stations and vice versa. Finally, the above reinforces the concept of a Mediterranean Oscillation between the western and eastern basins. Received November 14, 1997 Revised June 2, 1998     

Relationship of Late-Winter Temperatures in Europeto North Atlantic Surface Winds: A Correlation Analysis

Summary  The relationship between European surface temperature and winds over the eastern North Atlantic are investigated for the years 1988 to 1997. Daily Special Sensor Microwave Imager SSM/I observations are used to evaluate a monthly surface wind index that quantifies the influence of southwesterly flow. Our wind index and the monthly-mean surface-air temperatures in late winter and early spring over France and northern-latitude Europe are highly correlated. In February, the year-to-year increases/decreases match every year for France (correlation of 0.82 with the Index); and every year with just one exception for Europe (correlation with the Index of 0.76 for a longitudinal strip through Europe 45–50° N, and 0.73 for the 50–60° N strip). In March, the increases/decreases of the wind Index and of the temperatures for France also match, but the correlation with the Index is lower, 0.65. The high correlation between our Index and the large interannual fluctuations in the monthly temperature in late winter and early spring indicate that the onset of the spring conditions in Europe is significantly influenced by the wind patterns over the eastern North Atlantic. Coinciding with the fluctuations from warm-Europe/high-Index winter to the opposite conditions, we observe “seesaw” effects, fluctuations over the North Atlantic, in opposite directions in the east (25–5° W), and the west (65–45° W). In the low-Index years we find that: (a) the surface-air temperatures in the west are appreciably higher than in the east (but slightly lower in the high-Index year), and (b) the difference between the 500 mb meridional wind in the west and that in the east is positive and large, exceeding 10 m s⁻¹ (but it becomes negative and small in the high-Index years). The “seesaw” effects suggest that a positive feedback exits between these cross-Atlantic temperature differences and the surface winds. Received August 7, 1998 Revised April 23, 1999     

Interpretation of recent temperature and precipitationtrends observed in Korea

Summary  The possibility of climate change in the Korean Peninsula has been examined in view of the general increase in greenhouse gases. Analyses include changes in annual temperature and precipitation. These analyses are supplemented with our observations regarding the apparent decrease of forest areas. It was found that there was a 0.96 °C (0.42 °C per decade) increase in annual mean temperature between 1974 and 1997. The increase in large cities was 1.5 °C but only 0.58 °C at rural and marine stations. The difference in the mean temperature between large cities and rural stations was small from 1974 to 1981. However, the difference increased from 1982 to 1997. In particular, the warming appears most significant in winter. Prior to 1982, the lowest temperatures were often −18 °C in central Korea, and since then the lowest temperatures have been only −12∼−14 °C. Recently, the minimum January temperature has increased at a rate of 1.5 °C per decade. It is estimated that the increase of1 °C in annual mean temperature corresponds to about a 250 km northward shift of the subtropical zone boundary. The analysis of data from 1906 to 1997 indicates a trend of increasing annual precipitation, an increase of 182 mm during the 92-year peirod, with large year-to-year variations. More than half of the annual mean amount, 1,274 mm, occurred from June to September. Meteorological data and satellite observations suggest that changes have occurred in the characteristics of the quasi-stationary fronts that produce summer rain. In recent years scattered local heavy showers usually occur with an inactive showery front, in comparison with the classical steady rain for more than three weeks. For instance, local heavy rainfall, on 6 August 1998 was in the range of 123–481 mm. The scattered convective storms resulted in flooding with a heavy toll of approx. 500 people. The northward shift of the inactive showery front over Korea, and of a convergence zone in central China, correlate with the increase in temperature. It has been suggested that the decrease in forest areas and the change in ground cover also contribute to the warming of the Korean Peninsula. Received March 16, 2000     

Eastern Mediterranean summer temperatures since 730 CE from Mt. Smolikas tree-ring densities

The Mediterranean has been identified as particularly vulnerable to climate change, yet a high-resolution temperature reconstruction extending back into the Medieval Warm Period is still lacking. Here we present such a record from a high-elevation site on Mt. Smolikas in northern Greece, where some of Europe’s oldest trees provide evidence of warm season temperature variability back to 730 CE. The reconstruction is derived from 192 annually resolved, latewood density series from ancient living and relict Pinus heldreichii trees calibrating at r1911–2015 = 0.73 against regional July–September (JAS) temperatures. Although the recent 1985–2014 period was the warmest 30-year interval (JAS Twrt.1961–1990 = + 0.71 °C) since the eleventh century, temperatures during the ninth to tenth centuries were even warmer, including the warmest reconstructed 30-year period from 876–905 (+ 0.78 °C). These differences between warm periods are statistically insignificant though. Several distinct cold episodes punctuate the Little Ice Age, albeit the coldest 30-year period is centered during high medieval times from 997–1026 (− 1.63 °C). Comparison with reconstructions from the Alps and Scandinavia shows that a similar cold episode occurred in central Europe but was absent at northern latitudes. The reconstructions also reveal different millennial-scale temperature trends (NEur = − 0.73 °C/1000 years, CEur = − 0.13 °C, SEur = + 0.23 °C) potentially triggered by latitudinal changes in summer insolation due to orbital forcing. These features, the opposing millennial-scale temperature trends and the medieval multi-decadal cooling recorded in Central Europe and the Mediterranean, are not well captured in state-of-the-art climate model simulations.     

Monthly air temperature trends in Switzerland 1901–2000 and 1975–2004

Summary We analysed long-term temperature trends based on 12 homogenised series of monthly temperature data in Switzerland at elevations between 316 m.a.s.l. and 2490 m.a.s.l for the 20^th century (1901–2000) and for the last thirty years (1975–2004). Comparisons were made between these two periods, with changes standardised to decadal trends. Our results show mean decadal trends of +0.135 °C during the 20^th century and +0.57 °C based on the last three decades only. These trends are more than twice as high as the averaged temperature trends in the Northern Hemisphere. Most stations behave quite similarly, indicating that the increasing trends are linked to large-scale rather than local processes. Seasonal analyses show that the greatest temperature increase in the 1975–2004 period occurred during spring and summer whereas they were particularly weak in spring during the 20^th century. Recent temperature increases are as much related to increases in maximum temperatures as to increases in minimum temperature, a trend that was not apparent in the 1901–2000 period. The different seasonal warming rates may have important consequences for vegetation, natural disasters, human health, and energy consumption, amongst others. The strong increase in summer temperatures helps to explain the accelerated glacier retreat in the Alps since 1980. Authors’ addresses: Martine Rebetez, WSL Swiss Federal Research Institute, 1015 Lausanne, Switzerland; Michael Reinhard, Laboratory of Ecological Systems (ECOS), EPFL Swiss Federal Institute of Technology, 1015 Lausanne, Switzerland.     

利用采暖/制冷度日分析建筑能耗变化的适用性评估

采暖/制冷度日与建筑能耗有显著线性关系,被认为是最简单可靠的衡量能源需求的指标,但其适用性还缺乏全面的评估。通过模拟1961-2009年天津市办公、商场以及不同节能水平的居住建筑能耗,分析了能耗与采暖/制冷度日的关系,确定其反映能耗变化的适用性。结果表明：采暖期办公及商场建筑的热负荷与采暖度日的相关性达到极显著水平（P<0.01）,决定系数（R2）分别为0.99和0.97;制冷期冷负荷与制冷度日的关系尽管也达到极显著水平,但决定系数仅分别为0.64和0.55;不同节能水平居住建筑热负荷与采暖度日极显著相关,决定系数均在0.99以上。研究认为：采暖度日可以反映办公、商场及居住建筑的热负荷特征,用于分析气候变化对能耗的影响是可行的;但制冷度日不能完全反映办公及商场建筑冷负荷的变化,仅可分别解释冷负荷变化的64%和55%,单纯用制冷度日研究能耗变化是不够全面的。通过分析能耗与气候要素的关系发现,冬季采暖期能耗主要受气温的影响,而夏季制冷能耗受气温和湿度的共同影响。     

Last Glacial Maximum climate of the former Soviet Union and Mongolia reconstructed from pollen and plant macrofossil data

 An improved concept of the best analogues method was used to reconstruct the Last Glacial Maximum (LGM) climate from a set of botanical records from the former Soviet Union and Mongolia. Terrestrial pollen and macrofossil taxa were grouped into broad classes – plant functional types (PFTs), defined by the ecological and climatic parameters used in the BIOME1 model. PFT scores were then calibrated in terms of modern climate using 1245 surface pollen spectra from Eurasia and North America. In contrast to individual taxa, which exhibit great variability and may not be present in the palaeoassemblages, even in suitable climates, PFTs are more characteristic of the vegetation types. The modified method thus allows climate reconstruction at time intervals with partial direct analogues of modern vegetation (e.g. the LGM). At 18 kBP, mean temperatures were 20–29 °C colder than today in winter and 5–11 °C colder in summer in European Russia and Ukraine. Sites from western Georgia show negative, but moderate temperature anomalies compared to today: 8–11 °C in January and 5–7 °C in July. LGM winters were 7–15 °C colder and summers were 1–7 °C colder in Siberia and Mongolia. Annual precipitation sums were 50–750 mm lower than today across northern Eurasia, suggesting a weakening of the Atlantic and Pacific influences. Reconstructed drought index shows much drier LGM conditions in northern and mid-latitude Russia, but similar to or slightly wetter than today around the Black Sea and in Mongolia, suggesting compensation of precipitation losses by lower-than-present evaporation. Received: 11 May 1998 / Accepted: 25 September 1998     

Evaluation of urban thermal environments in commercial and residential spaces in Okayama City, Japan, using the wet-bulb globe temperature index

We estimated wet-bulb globe temperature (WBGT) using measured meteorological data to understand the bioclimates of human living spaces during the summer season. Our research focused on commercial and residential areas of Okayama City, Japan (population ～700,000). The commercial spaces (CO) mainly consisted of multi-story office buildings, whereas the residential spaces (RE) consisted of one- or two-story residential buildings. On a fine day with southeast winds, the spatially averaged WBGT measured in the CO was higher than that in the RE. The difference was statistically significant and would have caused noticeable discomfort and a high risk of heat disorder for occupants of the CO over the long term. For instance, at 1900 Japan Standard Time (JST), the maximum difference in the WBGT between the CO and RE sites was 2.0°C (23.5°C for the CO and 21.5°C for the RE). From 1800 to 1900 JST, the wet-bulb temperature in the CO was still 1.5–2.0°C higher than that in the RE, even though both areas had the same dry-bulb temperature. This indicates that the CO retained greater amounts of water vapor for longer periods compared to the RE. The wet-bulb temperature in the CO increased rapidly at most observation points when the southeast sea breeze arrived. In contrast, in the RE, the wet-bulb temperature decreased until evening. This difference was caused by moist air transported from a river about 1 km upwind from the CO. The moist air forced an increase in the WBGT and elevated the risk of heat disorder in the CO. The spatially averaged globe temperature of the CO at 1500 JST was 6.2°C lower than that at the RE, causing the WBGT of the CO to decrease. The results suggest that the higher WBGT in the CO was caused by higher wet-bulb temperatures. On a day with southwest winds, the CO and RE showed no difference in WBGT because the river was not included in the upwind source area.     

The Urban Heat Island Effect at Fairbanks, Alaska

Summary  Using climatic data from Fairbanks and rurally situated Eielson Air Force Base in Interior Alaska, the growth of the Fairbanks heat island was studied for the time period 1949 – 1997. The climate records were examined to distinguish between a general warming trend and the changes due to an increasing heat island effect. Over the 49-year period, the population of Fairbanks grew by more than 500%, while the population of Eielson remained relatively constant. The mean annual heat island observed at the Fairbanks International Airport grew by 0.4 °C, with the winter months experiencing a more significant increase of 1.0 °C. Primary focus was directed toward long-term heat island characterization based on season, wind speed, cloud cover, and time of day. In all cases, the minima temperatures were affected more than maxima and periods of calm or low wind speeds, clear winter sky conditions, and nighttime exhibited the largest heat island effects. Received August 17, 1998 Revised March 26, 1999     

Estimating the annual mean screen temperature empirically

Summary We have examined station data from around the world to study the separate effects of the latitude (between 60° N–40° S), elevation and distance inland, on the annual-mean screen temperature. In the first 200–400 km from some west coasts, screen temperatures (after adjustment for elevation) rise inland, reaching a maximum called the ‘thermal-ridge temperature’ Tr. The rise of temperature within this littoral fringe (of width F) depends mainly on the difference between the sea-surface temperature off the west coast and the zonal mean. Further inland than such a fringe, adjusted temperatures generally decline eastwards, approximately linearly, at a rate C. The rate is related to hemisphere and latitude. Empirical relationships between latitude and the observed coastal sea-surface temperature, the near-shore screen temperature, Tr, C and F for each continent are used to estimate annual mean temperatures on land. Independent estimates of this kind for 48 places, using a look-up table, differ overall by only 0.7 K from the actual long-term average annual mean temperatures. This is less than half the error resulting from an assumption of zonal-mean temperatures. Basing estimates on coastal sea-surface temperatures, instead of the look-up table, results in an average error of 1.0 K for the 48 places. The errors are comparable with the standard deviation of annual mean temperatures during 30 years or so. Received March 6, 2001 Revised July 30, 2001     

The semiannual oscillation and Antarctic climate, part 5: impact on the annual temperature cycle as derived from NCEP/NCAR re-analysis

We use NCEP/NCAR reanalysis data to study the impact of the semiannual oscillation (SAO) on the annual cycle of Antarctic near-surface temperature. When the SAO is weak, the contracted phases (March/April and September/October) are warm and the expanded phases (December/January and June/July) cold. This pattern is explained in terms of the changing meridional fetch of the circumpolar pressure trough. Because of the wave number three character of the SAO, large regional deviations are found. For instance, enhanced north-westerly flow in the second expansion phase (June/July) of weak SAO years limits the growth of the sea ice in the Amundsen and Bellingshausen seas, leading to anomalously high temperatures in the Antarctic Peninsula region. The short (<50 year) temperature records at Antarctic stations still carry the fingerprint of decadal SAO variability. By matching the observed monthly temperature trends to the patterns derived from the gridded re-analysis, we propose a background Antarctic warming trend for the second expansion phase (June/July) of 4.62 ± 1.02 °C per century, four times the annual value. Received: 23 August 1999 / Accepted: 28 October 1999