The social forecasting industry

This is an essay and report on the social forecasting industry-the collection of individuals and organizations, both profit and nonprofit, supplying forecasts of social futures in direct exchage for pay. It begins with a review of the industry's size, growth, and scope; contrary to initial impressions, the industry is by any comparative standard quite small, occupying a precarious niche in the much larger market for statistical services. Hallmark methodologies of the industry, including the Delphi method and cross-impact analysis, are discussed extensively. These methodologies are linked to a domain assumption of the industry—namely, that the complexities of the social world are best rendered tractable by forming consensuses of experts' judgments. It is argued that the key characteristic differentiating the social forecasting industry from forecasting in the social sciences proper—the existence of a direct, paying market for services—is an important element in maintaining the most salient methodological and product differences between the two arenas: The marketplace for forecasts is not the same as the marketplace for ideas.     

Recent developments in technological forecasting

Technological forecasting means the prediction of characteristics or use of technology. The methods used by technological forecasters are in principle no different from those used by forecasters in other application areas. However, the unique problems of the field require that the methods be adapted to those problems. This paper discusses recent developments involving refinements in the methods which have been in use for the past several decades. It also describes some important recent work on estimating upper limits to the progress of technologies, and on quantitative measures of multi-attribute technologies. Finally, it discusses several issues common to all forecasting application areas, as they are dealt with in technological forecasting. These issues include validation, disasters of forecasting, determinism in forecasting, and some examples of forecasts with practical applications.     

Some investigations in hydrological forecasting

Summary In the northern parts of Sweden precipitation falls during the winter as snow which melts only in spring, i. e. in April and May, or in the northern mountainous regions in June or even in July. Under natural conditions the water level and the flow are high during late spring and summer and low in winter. In order to reach a better utilization of the water power storage magazines have been created and thus the flow can be regulated in order to obtain more power during the winter time. To be able to do this in the most profitable way it is necessary to find the relations between the net influx and the relevant meteorological factors and then to make forecasts concerning the water supply for rather long periods. Some results from such already established forecast relations are discussed. There forecasts are concerned with the total flow of water during the spring time and are based only upon the amount of precipitation during the winter. Some preliminary results are also given from trials, just being started and aiming at forecasts of the variation with time of the flow during the spring.

---

Zusammenfassung Der Niederschlag in den n?rdlichen Teilen von Schweden f?llt w?hrend des Winterhalbjahres in Form von Schnee, der erst im Frühling, d. h. im April und Mai oder in den n?rdlichen Gebirgslagen sogar erst im Juni oder Juli, schmilzt. In den natürlichen Wasserl?ufen sind Wasserstand und Abflu? hoch w?hrend des Sp?tfrühlings und des Sommers, niedrig w?hrend des Winters. Um die Wasserkraft besser ausnutzen zu k?nnen, wurden Wasserreservoire geschaffen, wodurch die Wassermenge reguliert und die Energie auch im Winter ausgenützt werden kann. Um diese Wasserkraft in der günstigsten Weise ausnützen zu k?nnen, ist es notwendig, die Beziehungen zwischen dem Zuflu? und den wirksamen meteorologischen Faktoren festzulegen und nach M?glichkeit Prognosen über die Wasserausbeute auf l?ngere Sicht zu machen. Im vorliegenden Aufsatz werden die Ergebnisse von einigen solchen Prognoseversuchen er?rtert, die nur auf dem Winterniederschlag basieren und den totalen Abflu? ergeben. Weiterhin werden einige vorl?ufige Ergebnisse von neuerdings angefangenen Versuchen berichtet, die zeitlichen Schwankungen des Frühjahrshochwassers vorauszusagen. Résumé Dans les parties septentrionales de la Suède, les précipitations tombent, pendant la saison hivernale sous la forme de neige, qui ne se fond qu'au printemps, c'est-à-dire en avril et mai ou, dans les parties montagneuses du nord, même en juin ou juillet. Dans les courants d'eau naturels la hauteur d'eau et le débit sont élevés vers la fin du printemps et en été, bas pendant l'hiver. Pour mieux exploiter les forces hydrauliques, des réservoirs d'eau ont été construits, par lesquelles le débit peut être réglé et la force hydraulique peut être exploitée aussi pendant la saison d'hiver. Pour arriver à des résultats plus favorables il est nécessaire d'étudier les relations entre l'afflux et les facteurs météorologiques et de faire dans la mesure du possible des prévisions de l'évolution du débit d'eau à plus longue échéance. Quelques résultats de tels essais de prognostics établis, basés uniquement sur la précipitation d'hiver et donnant le volume total d'eau, sont discutés. En outre, on donne des résultats provisoires des essais qui ont été commencés récemment et qui concernent les possibilités de la prévision des variations temporaires des grandes crues de printemps.

---

---

With 2 Figures

---

Dedicated to Dr.Anders K. ?ngstr?m on the occasion of his 70th birthday.     

Modeling olive-crop forecasting in Tunisia

Tunisia is the world’s second largest olive oil-producing region after the European Union. This paper reports on the use of models to forecast local olive crops, using data for Tunisia’s five main olive-producing areas: Mornag, Jemmel, Menzel Mhiri, Chaal, and Zarzis. Airborne pollen counts were monitored over the period 1993–2011 using a Cour trap. Forecasting models were constructed using agricultural data (harvest size in tonnes of fruit/year) and data for several weather-related and phenoclimatic variables (rainfall, humidity, temperature, Growing Degree Days, and Chilling). Analysis of these data revealed that the amount of airborne pollen emitted over the pollen season as a whole (i.e., the Pollen Index) was the variable most influencing harvest size. Findings for all local models also indicated that the amount, timing, and distribution of rainfall (except during blooming) had a positive impact on final olive harvests. Air temperature also influenced final crop yield in three study provinces (Menzel Mhiri, Chaal, and Zarzis), but with varying consequences: in the model constructed for Chaal, cumulative maximum temperature from budbreak to start of flowering contributed positively to yield; in the Menzel Mhiri model, cumulative average temperatures during fruit development had a positive impact on output; in Zarzis, by contrast, cumulative maximum temperature during the period prior to flowering negatively influenced final crop yield. Data for agricultural and phenoclimatic variables can be used to construct valid models to predict annual variability in local olive-crop yields; here, models displayed an accuracy of 98, 93, 92, 91, and 88 % for Zarzis, Mornag, Jemmel, Chaal, and Menzel Mhiri, respectively.     

支持向量机在雷雨天气预报中的应用

支持向量机(SVM)是一种新型的机器学习方法。利用1999-2003年7月清远站每天08:00的探空资料,建立广州白云机场24 h内有无雷雨的SVM分类模型,进行相应的预报实验,实验结果显示对应的SVM分类模型效率高、准确率高,且泛化能力强,预报Ts评分非常理想,都达到80%以上;结果准确率并不会因为训练样本数目的减少而大幅度降低,具有良好的预报能力。对于某个特定的核函数,可通过调整误差惩罚参数C来得到性能最优的SVM。     

地面实况资料在暴雨预报中的运用

根据2005年的几次暴雨过程,探讨了地面实况资料与数值预报产品结合运用的问题。分析表明:四川盆地冷空气的活动,对贵州出现暴雨过程有重要影响,正确判断未来辐合区的位置对暴雨落区预报至关重要。     

Errors in wind-speed measurements by rotation anemometers

A perturbation theory approach of non-linear mechanics is applied to the solution of a non-linear rotation anemometer dynamic equation in a gusty wind. The first two terms of the perturbation series give a simplified equation for the wind-speed overestimation by a rotation anemometer (u-error) in terms of a wind velocity spectrum (or a correlation function). The equation agrees satisfactorily with all the known analytical or numerical solutions of rotation anemometer equations. It agrees, in particular, with recent theoretical estimations of the u-error magnitude by Kondo et al. (1971) and Hyson (1972), but disagrees significantly with the experimental findings of Izumi and Barad (1970) and Högström (1974). The same approach is also used for the estimation of the influence of the vertical wind fluctuations on the rotation anemometer readings (w-error). It is shown that w-error is usually of the same sign as u-error and that the sum of these both types of errors may be in some cases of the same order as an experimental wind-speed overestimation observed by Izumi and Barad and by Högström. However, it seems probable that some additional types of errors contribute also to the real overspeeding of rotation anemometers in a gusty wind.     

江西一次极端雷暴大风过程的中尺度特征与成因分析

2021年5月15日贵州东北部发生了一次极端风雹天气过程。本文利用常规观测资料以及加密自动气象站、探空、卫星、多普勒天气雷达等资料,分析了该过程的环境场和雷达回波特征。结果表明: (1) 此次极端风雹天气发生在低层暖湿和对流层中下层(500 hPa以下)具有显著条件不稳定的环境下,冷锋前沿的地面辐合线是其触发机制,切变线南侧暖湿气流输送和较强的垂直风切变有利于对流风暴的维持和加强。(2) 贵州东北部较大的温度递减率和对流有效位能,强的垂直风切变,较低的自由对流高度是冰雹和雷暴大风发生的有利条件。雷暴系统过境使受其影响的测站风速急增、气压猛升、气温骤降,表现为典型的强雷暴系统过境特征。(3) 雷暴系统由初始线状多单体逐渐形成分布不规则的多单体风暴系统。雷暴单体在反射率因子图上呈现出强回波中心超过60 dBz及其南侧高反射率因子梯度和中高层明显回波悬垂特征。造成下击暴流的雷暴单体的弓形回波特征明显,并伴有前侧入流缺口、后侧“V”型缺口及中层4—8 km高度强径向气流辐合(MARC)特征;而造成冰雹的雷暴单体的钩状回波特征十分明显,同时伴有较大垂直积分液态水含量(VIL)、较高回波顶高、强风暴顶辐散以及三体散射现象。

     

北京盛夏一次罕见的大雹事件分析

利用MICAPS、多普勒雷达资料和中尺度气象资料,统计了近9 a来金华地区共18次冰雹个例,并对其形成的天气学机制和雷达回波特征进行了分析研究。结果表明：（1）金华地区的冰雹天气型分为西风槽型和副热带高压影响型两类,年际高发时段集中在春季和盛夏,日际则集中在午后。（2）对流有效位能和风垂直切变是影响该区域冰雹产生的重要因子,夏季冰雹大多产生在高能弱切变环境下,春季冰雹大多产生在低能强切变环境下,大冰雹多出现在高能环境下,极易出现在高能且中等以上切变环境中。（3）冰雹发生前大气的水汽含量较高,整层大气可降水量（PWV）的平均值可达40.8 mm,而湿球温度0℃层则适宜。（4）冰雹单体生命史均超过1 h,0.5°仰角最大反射率因子值超过60 dBz;冰雹云风暴顶高分布与回波顶高度分布一致,且具有较明显的季节变化。（5）单体的垂直累积液态水含量（VIL）先增后降,中间至少有一次跃增过程,其最大值出现后的突降时间与降雹时间基本一致;强天气概率（SWP）产品相对冰雹出现有12~160 min不等的提前量。

     

榆林市地质灾害气象预报方法初探

统计分析1985—2007年榆林市地质灾害资料及对应的降水资料,初步揭示了榆林市地质灾害规律及地质灾害发生与降水的关系。结果表明:地质灾害与前10天降水量有明显的相关性。建立基于日综合雨量预测地质灾害的预报模型,确立了榆林地质灾害等级预报的降水指标,对榆林市地质灾害预报预警有一定的参考价值。     

对天气预报工作的思考

本文就天气预报工作性质、天气预报工作目标以及天气预报的管理工作．阐述一些认识和观点。     

对天气预报工作的思考

本文就天气预报工作性质、天气预报工作目标以及天气预报的管理工作,阐述一些认识和观点.     

冰雹云的多普勒天气雷达识别参量及其预警作用

利用呼和浩特多普勒天气雷达、卫星云图及NCEP (1°×1°)逐6 h再分析资料,对2015年7月29日呼和浩特市的一次冷涡后部环境条件下发生的冰雹天气过程进行分析,结果表明：在冷涡后部,高层冷平流和低层暖平流有利于对流不稳定度加大,较大的对流有效位能(CAPE)和假相当位温(θse)高能区反映了不稳定能量在冰雹区积聚;中低层水汽输送、0~6 km风场强烈顺时针旋转和中等强度的垂直风切变,为冰雹发生提供可能;地面辐合线、冷空气侵入及干线是冰雹的主要触发条件,冰雹主要发生在低层高能区和干线附近靠近湿区一侧。组合反射率因子图上呈现“钩状回波”和“弓形回波”,最强反射率因子大于等于60 dBz,反射率垂直剖面图上可识别低层弱回波区和中高层回波悬垂,回波顶高14~15 km,大于等于45 dBz的回波达到8 km,扩展到-20 ℃高度层;径向速度图上具有气旋式旋转辐合和“逆风区”;垂直累积液态水含量VIL达到64 kg·m^-2,且降雹前有明显的跃增现象。

     

MLP-based drought forecasting in different climatic regions

Water resources management is a complex task and is further compounded by droughts. This study applies a multilayer perceptron network optimized using Levenberg–Marquardt (MLP) training algorithm with a tangent sigmoid activation function to forecast quantitative values of standardized precipitation index (SPI) of drought at five synoptic stations in Iran. The study stations are located in different climatic regions based on De Martonne aridity index. In this study, running series of total precipitation corresponding to 3, 6, 9, 12, and 24?months were used and the corresponding SPIs were calculated: SPI3, SPI6, SPI9, SPI12, and SPI24. The multilayer perceptrons (MLPs) for SPIs with the 1-month lead time forecasting, were tested and validated. Four different input vectors were considered during network development. In the first model, MLP constructed by importing antecedent SPI with 1-, 2-, 3-, and 4-month time lags and antecedent precipitation with 1- and 2-month time lags (MLP1). Addition of antecedent North Atlantic Oscillation or antecedent Southern Oscillation Index with 1-month time lag or both of them to MLP1 led to MLP2, MLP3, and MLP4, respectively. The MLP models were evaluated using the root mean square error (RMSE) and the coefficient of determination (R ²). The results showed that MLP4 had a higher prediction efficiency than the other MLPs. The more satisfactory results of RMSE and R ² values of MLP4 for 1-month lead time for validation phase were equal to 0.35 and 0.92, respectively. Also, results indicated that MLPs can forecast SPI24 and SPI12 more accurately than the other SPIs.     

Errors in turbulence measurements with a sonic anemometer

Errors in surface-layer wind data obtained by sonic anemometers are estimated as a function of relative wind direction using a comparison between two identical anemometers. The maximum errors in mean windspeed and the standard deviations of the horizontal wind components are found to be between 10 and 20%. Agreement between the size of the measured errors and those calculated from wind-tunnel data is fair.In addition, vertical velocity data are analysed to show the presence of flow distortion induced by the asymmetry in the sonic anemometer probe and the mounting of the probe on the top of a mast.     

Errors and correction of precipitation measurements in China

In order to discover the range of various errors in Chinese precipitation measurements and seek a correction method, 30 precipitation evaluation stations were set up countrywide before 1993. All the stations are reference stations in China. To seek a correction method for wind-induced error, a precipitation correction instrument called the ＂horizontal precipitation gauge＂ was devised beforehand. Field intercomparison observations regarding 29,000 precipitation events have been conducted using one pit gauge, two elevated operational gauges and one horizontal gauge at the above 30 stations. The range of precipitation measurement errors in China is obtained by analysis of intercomparison measurement results. The distribution of random errors and systematic errors in precipitation measurements are studied in this paper. A correction method, especially for wind-induced errors, is developed. The results prove that a correlation of power function exists between the precipitation amount caught by the horizontal gauge and the absolute difference of observations implemented by the operational gauge and pit gauge. The correlation coefficient is 0.99. For operational observations, precipitation correction can be carried out only by parallel observation with a horizontal precipitation gauge. The precipitation accuracy after correction approaches that of the pit gauge. The correction method developed is simple and feasible.     

Severe convective storms and associated phenomena in Hungary

Convective activity dominates the weather of Hungary in the summer. Especially during the first part of the summer, the frequency of severe thunderstorms grows and associated phenomena such as wind storms, hail, sometimes even tornadoes cause serious damage. In this paper, an overview of the severe thunderstorm situation in the Carpathian Basin is presented with a focus on the most frequent phenomena: squall lines. To understand the physical background of these kinds of severe thunderstorms, the terminology of convective components is introduced. Making use of case studies, the roles of convective components are shown for different types of thunderstorm systems. Case studies also show that most of the tornado events are associated with organized thunderstorm systems, especially prefrontal squall lines.     

近地面层湍流通量观测误差的比较

本文根据误差理论分析并比较了在近地面层内各种间接确定湍流通量方法的观测误差。空气动力学法由于采用不同的通用函数将造成很大的误差。H扣除法确定潜热通量,当Bowen比较小时误差较小,Bowen比B>1之后误差随B的增大迅速增长。在相同的观测条件下间接法确定湍流通量以Bowen比法和组合法误差最小,实例计算表明组合法的精度最高。     

GRAPES_Meso背景误差特征及应用

基于2015年6月—2016年5月GRAPES_Meso有限区域中尺度数值预报模式产品,采用美国国家气象中心（NMC）方法和高斯函数拟合方案统计中国区域的背景误差和水平相关尺度随纬度、高度和季节的变化特征。结果表明:控制变量的背景误差与水平相关尺度不仅随高度和纬度有明显变化,其中非平衡Exner气压和比湿具有明显的局地性和季节变化特征。非平衡Exner气压的背景误差在青藏高原地区较大,且冬季最大,夏季最小。比湿背景误差在低纬度热带季风区较大,且夏季最大,冬季最小。非平衡Exner气压和比湿的水平相关尺度在冬季最大,夏季最小。同时文中采用随高度变化的水平相关尺度替换GRAPES-3DVar中单一尺度参数,1个月的分析和模式预报试验表明,6 h的位势高度预报在对流层有明显改进;风场分析及其12 h内的预报在平流层改进明显;对24 h不同量级降水的预报有显著正贡献,也显著改善24 h内的小雨、中雨和大雨的空报现象,明显改善12~24 h特大暴雨的漏报现象。