BNSClim meteorological part (version 2)

This is the Baltic and North Sea Climatology (BNSC) for the Baltic Sea and the North Sea in the range 47 ° N to 66 ° N and 15 ° W to 30 ° E. It is the follow-up project to the knsc climatology. The climatology was first made available to the public in March 2018 by ICDC and is published here in a slightly revised version 2. It contains the monthly averages of mean air pressure at sea level, and air temperature, and dew point temperature at 2 meter height. It is available on a 1 ° x 1 ° grid for the period from 1950 to 2015. For the calculation of the mean values, all available quality-controlled data of the DWD (German Meteorological Service) of ship observations and buoy measurements were taken into account during this period. Additional dew point values were calculated from relative humidity and air temperature if available. Climatologies were calculated for the WMO standard periods 1951-1980, 1961-1990, 1971-2000 and 1981-2010 (monthly mean values). As a prerequisite for the calculation of the 30-year-climatology, at least 25 out of 30 (five-sixths) valid monthly means to be present in the respective grid box. For the long-term climatology from 1950 to 2015, at least four-fifths valid monthly means had to be available. Two methods were used (in combination) to calculate the monthly averages, to account for the small number of measurements per grid box and their uneven spatial and temporal distribution: 1. For parameters with a detectable annual cycle in the data (air temperature, dew point temperature), a 2nd order polynomial was fitted to the data to reduce the variation within a month and reduce the uncertainty of the calculated averages. In addition, for the mean value of air temperature, the daily temperature cycle was removed from the data. In the case of air pressure, which has no annual cycle, in version 2 per month and grid box no data gaps longer than 14 days were allowed for the calculation of a monthly mean and standard deviation. This method differs from knsc and BNSC version 1, where mean and standard deviation were calculated from 6-day windows means. 2. If the number of observations fell below a certain threshold, which was 20 observations per grid box and month for the air temperature as well as for the dew point temperature, and 500 per box and month for the air pressure, data from the adjacent boxes was used for the calculation. The neighbouring boxes were used in two steps (the nearest 8 boxes, and if the number was still below the threshold, the next sourrounding 16 boxes) to calculate the mean value of the center box. Thus, the spatial resolution of the parameters is reduced at certain points and, instead of 1 ° x 1 °, if neighboring values are taken into account, data from an area of 5 ° x 5 ° can also be considered, which are then averaged into a grid box value. This was especially used for air pressure, where the 24 values of the neighboring boxes were included in the averaging for most grid boxes. The mean value, the number of measurements, the standard deviation and the number of grid boxes used to calculate the mean values are available as parameters in the products. The calculated monthly and annual means were allocated to the centers of the grid boxes: Latitudes: 47.5, 48.5,... Longitudes: —14.5, -13.5,... In order to remove any existing values over land, a land-sea mask was used, which is also provided in 1 ° x 1 ° resolution. In this version 2 of the BNSC, a slightly different database was used, than for the knsc, which resulted in small changes (less than 1 K) in the means and standard deviations of the 2-meter air temperature and dew point temperature. The changes in mean sea level pressure values and the associated standard deviations are in the range of a few hPa, compared to the knsc. The parameter names and units have been adjusted to meet the CF 1.6 standard.本数据集为波罗的海与北海气候数据集（Baltic and North Sea Climatology, BNSC），覆盖北纬47°至66°、西经15°至东经30°范围内的波罗的海与北海海域，是knsc气候数据集的后续项目。该气候数据集于2018年3月由ICDC首次向公众发布，本次发布的是经过小幅修订的第二版。数据集包含海平面平均气压、2米高度处的气温与露点温度的月平均值，空间分辨率为1°×1°，时间跨度为1950年至2015年。 在平均值计算过程中，本数据集纳入了德国气象局（German Meteorological Service, DWD）在该时段内所有经过质量控制的船舶观测与浮标测量数据。若有相对湿度与气温数据，可额外推算得到露点温度值。本数据集以世界气象组织（World Meteorological Organization, WMO）的标准时段1951-1980、1961-1990、1971-2000及1981-2010为基准，计算了各时段的月平均值。对于30年尺度的气候态计算，要求对应网格盒内至少需包含30个月份均值中的25个（占比不低于5/6）；而针对1950-2015年的长期气候态，则要求至少有4/5的月份均值为有效数据。 针对每个网格盒的观测数量稀少且时空分布不均的问题，本数据集采用两种结合的方法计算月平均值： 1. 对于数据中存在可检测年周期的参数（气温、露点温度），对数据拟合二阶多项式，以降低月内变异性并降低计算平均值的不确定性。此外，针对气温平均值，需去除数据中的日温度周期信号。对于无年周期的气压参数，第二版中要求每个月与网格盒内的有效数据缺口不得超过14天，方可计算月平均值与标准差。该方法与knsc及BNSC第一版不同，后者通过6日滑动窗口均值计算平均值与标准差。 2. 当观测数量低于特定阈值时（气温与露点温度的阈值为每个网格盒每月20次观测，气压为每个网格盒每月500次观测），将使用相邻网格盒的数据参与计算。邻域数据的使用分为两步：首先调用最近的8个邻盒数据，若仍未达到阈值，则进一步纳入外围的16个邻盒数据，以计算中心网格盒的平均值。此举会在部分区域降低参数的空间分辨率：当引入邻域数据时，原本1°×1°的网格会扩展至覆盖5°×5°的区域，再将其平均为网格盒值。该方法在气压计算中尤为常用，多数网格盒的平均值会纳入24个邻盒的数据。 数据集产品中包含平均值、观测次数、标准差以及用于计算平均值的网格盒数量等参数。 计算得到的月均与年均值均分配至网格盒中心，其坐标为：纬度：47.5、48.5……；经度：-14.5、-13.5……。为剔除陆地区域的无效值，本数据集使用了1°×1°分辨率的海陆掩膜数据，该掩膜文件亦随数据集一同提供。 相较于knsc气候数据集，BNSC第二版采用了略有差异的数据库，导致2米气温与露点温度的平均值及标准差出现小幅变化（变化量小于1开尔文）；海平面气压平均值及对应标准差的变化量约为数百帕。数据集的参数名称与单位已调整以符合CF 1.6标准。