A 217-year precipitation reconstruction in the Habahe area, Xinjiang, Northeast China

Evaluating long-term changes in precipitation resources is important for accurate hydrological evaluation and forecasting,water security and rational allocation of water resources.For this purpose in the Xinjiang Habahe area,tree-ring specimens were collected from Picea obo-vata,Larix sibirica,and Betula platyphylla to establish a tree-ring width chronology,which was used to analyse a correlation with the average temperature and precipitation per month for 1958-2016.Based on correlation coefficients for monthly temperature and precipitation with the chro-nology of tree-ring widths,radial tree growth was mainly restricted by precipitation,and tree-ring width chronology was significantly correlated with overall precipitation from the previous July to the next June(r=0.641,P<0.01).The above results were used to establish a transformation equa-tion,and the overall precipitation from the previous July to the following June from 1800 to 2016 in Habahe was reconstructed after adjusted degrees of freedom,and obtain an explanatory rate of the variation up to 41.1%(40.0%).In addition to the reliability of the reconstructed values,the stability of the conversion function was determined via the“leave-one-out”method,which is commonly used in research on tree rings,and by cross-checking the conversion function with the reduced error value(RE),product mean test(t),with a sign test(ST).During the last 217 years,there were nine dry periods:1803-1829,1861-1865,1872-1885,1892-1905,1916-1923,1943-1954,1961-1966,1973-1981,and 2005-2011;and 12 wet periods:1830-1834,1836-1860,1866-1871,1886-1891,1906-1915,1925-1930,1934-1942,1955-1960,1967-1972,1982-1996,2000-2004,and 2012-2016.Comparisons of the reconstruc-tions for neighboring regions and a spatial correlation analy-sis showed that the reconstructed sequence of the present precipitation data better represented the changes in precipi-tation in Habahe.Additionally,a power spectrum analysis revealed that precipitation over the past 217 years in Habahe Province exhibited 2-5 years of quasiperiodic variation.A power spectrum analysis and wavelet analysis indicated that El Niño-Southern Oscillation influenced the precipitation cycles.This reconstruction provides more information on high-frequency precipitation,which is an important supple-ment to the existing tree-ring reconstruction of precipitation in the study area.The reconstruction of regional high-resolu-tion precipitation changes over the last several hundred years provides unique,important data for understanding regional differences in climate at the decadal-centennial scale.

Evolution of Meteorological Conditions during a Heavy Air Pollution Event under the Influence of Shallow Foehn in Urumqi, China

The air pollution in Urumqi which is located on the northern slope of the Tianshan Mountains in northwestern China,is very serious in winter.Of particular importance is the influence of terrain-induced shallow foehn,known locally as elevated southeasterly gale(ESEG).It usually modulates atmospheric boundary layer structure and wind field patterns and produces favorable meteorological conditions conducive to hazardous air pollution.During 2013-17,Urumqi had an average of 50 d yr-1 of heavy pollution(daily average PM2.5 concentration>150μg m-3),of which 41 days were in winter.The majority(71.4%)of heavy pollution processes were associated with the shallow foehn.Based on microwave radiometer,wind profiler,and surface observations,the surface meteorological fields and boundary layer evolution during the worst pollution episode in Urumqi during 16-23 February 2013 are investigated.The results illustrate the significant role of shallow foehn in the building,strengthening,and collapsing of temperature inversions.There were four wind field patterns corresponding to four different phases during the whole pollution event.The most serious pollution phase featured shallow foehn activity in the south of Urumqi city and the appearance of an intense inversion layer below 600 m.Intense convergence caused by foehn and mountain-valley winds was sustained during most of the phase,resulting in pollutants sinking downward to the lower boundary layer and accumulating around urban area.The key indicators of such events identified in this study are highly correlated to particulate matter concentrations and could be used to predict heavy pollution episodes in the feature.

Effects of irrigation on precipitation in the arid regions of Xinjiang,China

Soil moisture is an important parameter for the interaction between soil and atmosphere. It is the sec- ond important factor that influences climate change, next to sea surface temperature （SST）. Most previous studies focused on the monsoon regions in East China, and only a few laid emphases on arid environments. In Xinjiang, which is located in Northwest China, the climate is typically arid and semi-arid. During the past 20 years, the pre- cipitation in Xinjiang has shown a significant increasing trend, and it is closely related to oasis irrigation. This paper aims at discussing whether abnormal soil moisture in spring can be the signal to forecast summer precipitation. The effects of abnormal soil moisture due to farm irrigation in spring in arid environments on regional climate are inves- tigated by using a regional climate model （RegCM3）. The results indicate that positive soil moisture anomaly in irrigated cropland surface in May led to an increase in precipitation in spring as well as across the whole summer. The impact could last for about four months. The effects of soil moisture on the surface air temperature showed a time-lagging trend. The summer air temperature declined by a maximum amplitude of 0.8℃. The increased soil moisture could enhance evaporation and ascending motion in the low troposphere, which brought in more precipi- tation. The soil moisture affected regional weather and climate mainly by altering the surface sensible and latent heat fluxes.

Diagnosis and Analysis of a Rainstorm Process in Summer in Hami,Xinjiang

Based on conventional meteorological detection data and NCEP global reanalysis data,a rainstorm process in Hami,Xinjiang at the end of July 2018 was diagnosed and analyzed.The results show that this rare rainstorm occurred in the large-scale circulation situation of South Asia high,two ridges and one trough.The heavy precipitation area appeared on the right side of the southeasterly wind convergence area at 700 hPa,southerly flow at 500 hPa and the inlet area of southwest jet core at 200 hPa.The structure of high-level divergence and low-and middle-level convergence and strong vertical speed were important dynamic conditions for the occurrence and development of this rare rainstorm.The warm and moist air from the southern Sea of Japan and the Bay of Bengal was transported through different routes,providing a constant supply of water vapor and energy to the rainstorm area.Both EC and NCEP models have good predictive significance for this rainstorm,and the former is slightly better than the latter in terms of prediction effect.

Validation of AIRS-Retrieved atmospheric temperature data over the Taklimakan Desert

The Taklimakan Desert, the world's second largest desert, plays an important role in regional climate change. Previousstudies on its spatial temperature features suffered from sparse conventional detection data, but the Atmospheric InfraredSounder (AIRS) provides excellent temperature retrievals with high spatiotemporal resolution. Validation of AIRS temperatureretrievals over desert regions with high land-surface emissivity, the key contributor to inversion error, is essential beforeusing these data in regional weather/climate modeling. This paper examines the correlation coefficients, root meansquare error (RMSE) and mean BIAS between AIRS-retrieved atmospheric temperature data and radiosonde observations(RAOBs) in the Taklimakan Desert hinterland and oases in the morning and at dusk. Firstly, the AIRS retrievals are consistentwith RAOBs and are more consistent in the morning than at dusk. The consistency is better over a small-scale desert oasisthan over a large-scale oasis in the morning and exhibits the opposite trend at dusk. The correlation coefficient over thehinterland is high in the morning but negative at dusk due to high desert-surface emissivity. Second, the RMSEs, which areall smaller than 3 K, are generally higher over desert sites than over oasis sites and slightly lower over a small-scale oasisthan over a large-scale oasis in the morning. At dusk, the RMSEs are higher over desert sites than over oases and slightlyhigher over a small-scale oasis than over a large-scale oasis. Furthermore, the RMSEs are generally higher in the morningthan at dusk over a large-scale oasis and lower in the morning than at dusk over a small-scale oasis. Third, the absolutemean BIAS values are mostly lower than 1 K. In the morning, relative to RAOB temperatures, the retrieval temperaturesare higher over desert sites but lower over oasis sites. At dusk, the retrieval temperatures are lower than RAOB temperaturesover both desert and oasis sites. The retrieval temperatures are higher than RAOB temperatures over desert sites inthe morning but slightly lower at dusk. Most absolute mean BIAS values are higher in the morning than at dusk overboth oasis and desert sites. Finally, the consistency between the AIRS and RAOB temperature data is high from 700 hPato 100 hPa in the morning and from 700 hPa to 300 hPa at dusk. The difference between the AIRS and RAOB temperaturedata is generally higher in the morning than that at dusk. The RMSE differences between the AIRS and RAOB data areslightly lower in the morning than at dusk and are lower in the middle layers between 700 hPa and 150 hPa than in the layersabove 150 hPa during both the morning and night. The BIAS is lower in the morning than at dusk below 300 hPabut higher in the upper layers. Moreover, the BIAS value is positive in the middle layers between 500 hPa and 150 hPaand negative at other levels at both times. Generally, the AIRS retrieval temperatures are reliable and can be used in furtherstudies in the Taklimakan Desert.

THE INFORMATION FORMULA OF NORMAL EQUAL CORRELATION FACTOR

Let X and Y be random vectors and m and n their dimensions respectively. When the joint probability density function of X and Y is a normal m+n-dimensional probability density function, the information I_Y(X) of random vector Y about random vector X