Kendall's Statistic Research Articles

Vegetation Dynamics in the Upper Guinean Forest Region of West Africa from 2001 to 2015

The Upper Guinea Forest (UGF) region of West Africa is one of the most climatically marginal and human-impacted tropical forest regions in the world. Research on the patterns and drivers of vegetation change is critical for developing strategies to sustain ecosystem services in the region and to understand how climate and land use change will affect other tropical forests around the globe. We compared six spectral indices calculated from the 2001–2015 MODIS optical-infrared reflectance data with manually-interpreted measurements of woody vegetation cover from high resolution imagery. The tasseled cap wetness (TCW) index was found to have the strongest association with woody vegetation cover, whereas greenness indices, such as the enhanced vegetation index (EVI), had relatively weak associations with woody cover. Trends in woody vegetation cover measured with the TCW index were analyzed using Mann–Kendall statistics and were contrasted with trends in vegetation greenness measured with EVI. In the drier West Sudanian Savanna and Guinean Forest-Savanna Mosaic ecoregions, EVI trends were primarily positive, and TCW trends were primarily negative, suggesting that woody vegetation cover was decreasing, while herbaceous vegetation cover is increasing. In the wettest tropical forests in the Western Guinean Lowland Forest ecoregion, declining trends in both TCW and EVI were indicative of widespread forest degradation resulting from human activities. Across all ecoregions, declines in woody cover were less prevalent in protected areas where human activities were restricted. Multiple lines of evidence suggested that human land use and resource extraction, rather than climate trends or short-term climatic anomalies, were the predominant drivers of recent vegetation change in the UGF region of West Africa.

Geo-spatial analysis of temporal trends of temperature and its extremes over India using daily gridded (1°×1°) temperature data of 1969–2005

Daily gridded (1°×1°) temperature data (1969–2005) were used to detect spatial patterns of temporal trends of maximum and minimum temperature (monthly and seasonal), growing degree days (GDDs) over the crop-growing season (kharif, rabi, and zaid) and annual frequencies of temperature extremes over India. The direction and magnitude of trends, at each grid level, were estimated using the Mann–Kendall statistics (α = 0.05) and further assessed at the homogeneous temperature regions using a field significance test (α=0.05). General warming trends were observed over India with considerable variations in direction and magnitude over space and time. The spatial extent and the magnitude of the increasing trends of minimum temperature (0.02–0.04 °C year−1) were found to be higher than that of maximum temperature (0.01–0.02 °C year−1) during winter and pre-monsoon seasons. Significant negative trends of minimum temperature were found over eastern India during the monsoon months. Such trends were also observed for the maximum temperature over northern and eastern parts, particularly in the winter month of January. The general warming patterns also changed the thermal environment of the crop-growing season causing significant increase in GDDs during kharif and rabi seasons across India. The warming climate has also caused significant increase in occurrences of hot extremes such as hot days and hot nights, and significant decrease in cold extremes such as cold days and cold nights.

Century scale climate change in the central highlands of Sri Lanka

In this study, an analysis of century scale climate trends in the central highlands of Sri Lanka is presented. Monthly rainfall and temperature records of the period 1869–2006 from five climatological stations were analyzed. The trend is calculated by the least square regression analysis and the significance of the observed trend is estimated using the Mann–Kendall statistic. The results clearly show that there is a statistically significant decrease in annual rainfall in the western slopes of the central highlands. Throughout the last century, the annual reduction of rainfall in Nuwara Eliya which is at an altitude of 1895 m was 5.2 mm/year. The decrease is largely due to the reduction in southwest monsoon rainfall which contributes to 75% of the total reduction. No significant change was observed on the eastern side of the central highlands which receives rainfall predominantly from the northeast monsoons. The mean annual temperature in the mountainous region shows a uniform increasing trend which is in line with the 100-year global temperature increase of 0.8 ± 0.2∘C. Kandy, which is at an altitude of 477 m and closely linked with the rainfall climatology of Nuwara Eliya, showed no significant change in the mean annual temperature. If the current trend continues, in another 100 years, western and eastern slopes of central highlands will receive the same amount of rainfall from the southwest monsoon and the northeast monsoon which will have far reaching consequences for Sri Lanka’s economy and the ecology of the hill country.

Runoff Trends Driven by Climate and Afforestation in a Pyrenean Basin

AbstractThe abandonment of traditional rural life in mountain areas has favoured the expansion of forest in the headwaters of Pyrenean rivers. In this paper, we (i) analyse hydro‐climatic trends at the annual and monthly scales in three nested sub‐catchments in a central Pyrenean basin and (ii) quantify the relative contribution of climate change and forest cover on the observed changes in runoff. Land use maps indicate an increase in the forest cover in all sub‐basins for the period 1987–2009. Non‐parametric Mann–Kendall statistic on annual and monthly hydro‐climatic data from 1965 to 2009 was used to identify trends in climate and runoff data. To assess the potential influence of the period length, trends were also examined for the 1941–2009 stretch. Upward trends were detected for temperature and potential evapotranspiration, particularly during summer and winter months. Precipitation trends and their significance depended on the length of the period considered, although overall results indicated a decrease, particularly for February and July. The effects of climate change and land cover change on annual runoff were identified by means of a change point analysis. Results indicated that a change in annual runoff took place in the 1980s, suggesting the non‐linearity in the relationship between precipitation and runoff, and thus the influence of land use factors. The influence of forest cover was confirmed by comparing observed runoff trends with those resulting from a hydro‐climatic model that does not take into account land use changes. Divergence between both trends revealed that forest cover can account for ~40% of the observed decrease in annual runoff. Copyright © 2015 John Wiley & Sons, Ltd.

Spatio-temporal variability of rainfall regime in the Brahmaputra valley of North East India

Monthly rainfall data, spanning over 110 years (1901–2010), were utilized for trend analysis at different spatial and temporal scales over the Brahmaputra valley, India. The Mann–Kendall statistic and Sen’s slope model were used to identify the trends and estimate the magnitude of change, respectively. Statistical significance of the decadal shifts in rainfall from the overall mean was estimated by using Cramer’s test. The analysis revealed decrease in annual as well as monsoon rainfall in the Brahmaputra valley during the last 110 years with large spatial and temporal variations. These decreasing trends of rainfall in the eastern part of the valley were statistically significant. Significant decreasing trend of monsoon rainfall during the recent 30-year period was due to significant decrease of July and September rainfall, and this trend was found to be consistent at different spatial scales. In the last decade (2001–2010) in particular, monsoon rainfall exhibited significant negative deviation from the normal due to three deficient years and absence of excess rainfall years. On the contrary, contribution of pre-monsoon and post-monsoon rainfall to annual total in the Brahmaputra valley increased during the recent 30-year period. Winter rainfall in the valley decreased during the last 30 years due to significant decrease of December rainfall in the eastern and central parts.

Trend Detection of Drought in Arid and Semi-Arid Regions of Iran Based on Implementation of Reconnaissance Drought Index (RDI) and Application of Non-Parametrical Statistical Method

Drought is known as one of the main natural hazards especially in arid and semi-arid regions where there are considerable issues in regard to water resources management. Also, climate changes has been introduced as a global concern and therefore, under conditions of climate change and global warming, the investigation of drought severity trend in regions such as Iran which is mainly covered by arid and semi-arid climate conditions is in the primary of importance. Therefore, in this study, based on the application of Reconnaissance Drought Index (RDI) for assessment drought severities, and also the implementation of non-parametric Mann- Kendall statistics and Sen’s slope estimator, the trends in different time series of RDI (3, 6, 9, 12, 18 and 24 monthly time series) were investigated. Results indicated the frequent decreasing trends in RDI time series particularly for long term time series (12, 18 and 24 monthly time series) than short term ones. Decreasing trend in RDI time series means the increasing trend in drought severities. Since the water resources especially ground water in most cases are affected by long term droughts, therefore, increasing trend in drought intensities in long term ones can be a threat for water resources management in surveyed areas.

Omega-3 polyunsaturated fatty acids and left ventricular diastolic dysfunction in postmenopausal women

Purpose: To asses fatty acids composition of erythrocyte membrane lipids in postmenopausal women depending on left ventricular (LV) filling parameters. Methods: 203 postmenopausal women were included in the investigation: 142 women with surgical menopause (average age 44,3±5,4 years, menopause duration 3,8±2,4 years); and 61 women with natural menopause (average age 46,2±4,7 years, menopause duration 4,3±2,1 years). All patients underwent general clinical examination, echocardiographic examination with systolic and diastolic LV function assessment. Besides that we assessed the proportion of different fatty acids in lipids of erythrocyte membranes. Control group consisted of 27 healthy menstruating women (mean age 39,2±4,2). Statistical methods such as Cruskell-Walles criteria and τ Kendall statistics were used. Results: 117 (57.6%) postmenopausal women had LV diastolic function disturbance (1st group) and 86 women had no (2nd group). Nobody of included postmenopausal women had LV systolic function failure. All patients in menopause had significant changes in fatty acids composition of erythrocyte membrane lipids in comparison with healthy women: saturated acids relative content was increased, and unsaturated acids relative content was decreased, especially in women with LV diastolic dysfunction. The most serious changes concerned omega-3 polyunsaturated acids: relative contents of alpha-linoleic acid in patients of the 1st group was 3,1 times less than in control group (P<0.001); patients of the 2nd group demonstrated 2,7 times decrease of this index in comparison with healthy women (P<0.001). The content of eicosapentaenoic acid in the 1st and the 2nd groups was 1,7 and 1,4 times less than in control group, respectively (P<0.01). The same changes in the 1st and the 2nd groups concerned to docosapentaenoic and arachidonic acids: they were 2,1 – 3 times less and 2,4 – 2,8 times less than in healthy women, respectively (P<0.001). The presence of LV diastolic dysfunction correlated negatively with alpha-linoleic, eicosapentaenoic acid content and common polyunsaturated acids content (τ from –0.32 to –0.58, P<0.001) and positively – with saturated acids relative content (τ = 0.55, P<0.001). Conclusions: Thus, disturbances in fatty acids composition of erythrocyte membrane lipids may play role in left ventricular dysfunction formation in postmenopausal women.

Impacts of climate change and land use change on runoff of forest catchment in northeast China

Hydrological processes change from the impacts of climate variability and human activities. Runoff in the upper reaches of the Hun-Taizi River basin, which is mainly covered by forests in northeast China, decreased from 1960 to 2006. The data used in this study were based on runoff records from six hydrological stations in the upper reaches of the Hun-Taizi River basin. Nonparametric Mann–Kendall statistic was used to identify change trends and abrupt change points and consequently analyze the change characteristics in hydrological processes. The abrupt change in the annual runoff in most subcatchments appeared after 1975. Finally, the effects of climate change and land cover change on water resources were identified using regression analysis and a hydrology model. Results of the regression analysis suggest that the correlation coefficients between precipitation and runoff prior to the abrupt change were higher compared with those after the abrupt change. Moreover, using hydrology model analysis, the water yield was found to increase because of the decrease in forest land. Copyright © 2012 John Wiley & Sons, Ltd.

Drying climate in Ghana over the period 1960–2005: evidence from the resampling-based Mann-Kendall test at local and regional levels

Trends in rainfall series were investigated at 16 stations in Ghana over the period 1960–2005. Time series were first de-correlated using an effective pre-whitening methodology and then submitted to the resampling-based Mann-Kendall test. Field significances were assessed using the regional average Kendall statistic. Although no significant changes were observed in annual rainfall, the analysis reveals: (a) a reduction in the number of wet season days totalling less than 20 mm of rainfall, between latitudes 6° and 9.5°N; (b) a delay (about 0.5 d year‐1) in the wet season onset at several locations throughout the country; and (c) a lengthening (about 0.1 d year‐1) of rainless periods during the wet season in the south and centre of Ghana. All these changes, which remained insignificant at more than half of the individual stations, were found to be regionally significant at the 95% confidence level. The results highlight the importance of evaluating regional significance when investigating climate trends. Editor Z.W. Kundzewicz Citation Lacombe, G., McCartney, M., and Forkuor, G., 2012. Drying climate in Ghana over the period 1960–2005: evidence from the resampling-based Mann-Kendall test at local and regional levels. Hydrological Sciences Journal, 57 (8), 1594–1609.

Analysis of impacts of climate variability and human activity on streamflow for a river basin in northeast China

► Long-term records of hydrological data showed the temporal variation of runoff. ► Non-parametric Mann–Kendall statistic and Yamamoto test method were used. ► The trends and the change point of the annual streamflow were identified. ► Study catchment is Hun River basin and Taizi River basin in the northeast of China. ► Streamflow records of six hydrological stations from 1961 to 2006 are used. Hydrological processes in river systems have been changing under the impacts of both climate variability and human activities. The non-parametric Mann–Kendall statistic was used to identify change trends and points in the annual streamflow in the Hun–Tai River basin in northeast China. The identifications were based on streamflow records from six hydrological stations during 1961–2006, and the purpose was to analyze the change characteristics of the hydrological processes. The results indicated that all hydrological stations presented downward trends in annual streamflows. Abrupt changes in the annual streamflow occurred around 1978 in the Hun River basin, and around 1998 in the Taizi River basin. The impact of climate variability on the mean annual streamflow was also analyzed based on the relationships among streamflow, precipitation, and potential evapotranspiration. Precipitation and meteorological data from 22 rainfall stations and 10 weather stations within the basin were employed in the analysis. Daily potential evapotranspiration was calculated using the Penman–Monteith equation. Climate variability was estimated to account for 43% of the reduction in the annual streamflow, and human activities accounted for about 57%.

Global trends in NDVI-derived parameters obtained from GIMMS data

The Normalized Difference Vegetation Index (NDVI) has been proven to be useful to assess vegetation changes around the world, in spite of limitations such as sensitivity to cloud or snow contamination. In order to map vegetation changes at global scale, this study uses NDVI time series provided by the GIMMS (Global Inventory Modeling and Mapping Studies) group, which were fitted annually to a double logistic function. This fitting procedure allowed for retrieval of NDVI-derived parameters which were tested for trends using Mann–Kendall statistics. These trends were validated by comparison at 73 ground control points documented as change hotspots. The obtained trends for NDVI-derived parameters provide information to the remote sensing community on the regions more susceptible to having suffered changes, to complement other traditional methods such as national inventories and field work. Additionally, the validation procedure can be applied to any results obtained from a global dataset.

On the relationship between the premonsoonal rainfall climatology and urban land cover dynamics in Kolkata city, India

AbstractEmpirical and modelling studies show that urbanization can have an impact on the environment. Relatively few studies have investigated urban effects on precipitation in India or other developing countries experiencing rapid urbanization. Furthermore, most precipitation‐related studies for India focus on monsoonal rainfall. However, premonsoonal periods (March–May) account for 12–14% of the annual cumulative rainfall in eastern India. The majority of premonsoonal rainfall (PMR) is convective and caused by mesoscale forcing, which may include urban effects. In this study, the area under scrutiny is a large urban area in eastern India, Kolkata city. Herein, our goal was to (1) produce a comprehensive characterization of historical land cover dynamics associated with the Kolkata megalopolis, (2) provide a spatio‐temporal climatology of PMR in the Kolkata region, and (3) identify possible associations between Kolkata's land cover and PMR. The analysis shows that the rate of change of urban land cover has increased by 50% compared to the period prior to India's independence in 1947. A multi‐scalar time series analysis with Mann–Kendall statistics indicated statistically significant increasing trends in rainfall over the last 50 years for two Kolkata stations and a nearby downwind station. Furthermore, there was no significant trend for cumulative PMR in less urbanized stations, the country of India, or the East Gangetic region. This finding suggests that the anomaly of the three stations, showing increasing trends in PMR, could be the effect of urban land cover change. Copyright © 2011 Royal Meteorological Society

Trends in extreme rainfall events of Bangladesh

A study of the variability of the extreme rainfall events in Bangladesh during the time period 1958–2007 has been carried out in this paper. Quality-controlled homogeneous daily precipitation records of nine stations distributed over Bangladesh are used for the study. A total of 15 annual and seasonal indices of rainfall are examined. Variability of annual and seasonal rainfall trends is also assessed. The Mann–Kendall statistic and Sen's Slope model are used to reveal the trends and estimate the magnitude of change, respectively. A significant increase of annual and pre-monsoon rainfall in Bangladesh is observed. In general, an increasing trend in heavy precipitation days and decreasing trends in consecutive dry days are observed. Significant change in most of the extreme rainfall indices are observed in Northwest Bangladesh.

Robustness and monotonicity properties of generalized correlation coefficients

A new class of generalized correlation coefficients that contains the Pearson and Kendall statistics as special cases was defined by Chinchilli et al. (2005) and applied to the estimation of correlations coefficients within the context of 2×2 cross-over designs for clinical trials. In this paper, we determine the infinitesimal robustness and local stability properties of these generalized correlation coefficients by deriving their corresponding influence functions. For cases in which the population distribution is a bivariate normal or a mixture of bivariate normal distributions we obtain explicit formulas, and establish monotonicity and sign-reverse rule properties of the generalized correlation coefficients.

The Occurrence and Persistence of MTBE in Groundwater in Windham, Maine, USA

A study was conducted from July 1998 through November 2007 on the occurrence and distribution of the fuel oxygenate methyl tert‐butyl ether (MTBE) in a large sand and gravel aquifer located in southern Maine. MTBE was detected in 44% of 129 water samples collected from monitoring wells in concentrations up to 38.7 µg/L (reporting limit = 0.1 µg/L). The number of wells with detectable quantities of MTBE declined slightly between 1999 and 2007, but in general MTBE persisted throughout the period of study. Overall, MTBE was detected more frequently in the shallow and more transmissive parts of the aquifer. There was a statistically significant difference (p &lt; 0.001) for MTBE concentrations relative to nearby land uses. MTBE was detected in 83% of the samples collected from wells in low‐density residential areas, in 50% of samples from urban areas, and in 60% of samples from undeveloped areas. The concentrations of MTBE in the test wells were compared across the sample dates for trends and seven wells had a positive trend (Mann–Kendall statistic), but none was significant at p &lt; 0.05. Nine wells had a negative trend, but only one was significant at p &lt; 0.05. Three wells had no trend. The absence of strong or even consistent trends indicates that MTBE persists in shallow groundwater, even after gasoline formulations were changed to reduce or eliminate MTBE.

Analysis of a 27‐year rainfall data (1977–2003) in the Sultanate of Oman

AbstractAnalysis of variation of rainfall in space, time and amounts, and its attendant effect on the ecosystem is vital in arid and semi‐arid environments where the resource is scare, highly variable and unpredictable. In this study, we analyse the characteristics of rainfall in the Sultanate of Oman using data recorded between 1977 and 2003. The data is divided into six geomorphic compartments to represent the various topographic regions in Oman. The average yearly rainfall varies from a low of 76.9 mm in the interior region to a high of 181.9 mm in the Dhofar Mountains, with an average of 117.4 mm for the whole country. Mann–Kendall statistics show a negative but insignificant rainfall trends for the datasets. In northern Oman, the main rainfall season occurs between December and April and that accounts for 57.8–82.9% of the annual rainfall. February and March record the highest rainfall accounting for 35.3 to 42% of the yearly rainfall. The Dhofar Mountains and surrounding areas in southern Oman are dominated by the khareef season in July–August, which produces 44.3 to 67.5% of the rainfall in that area. The number of days of light rainfall (&lt;10 mm per day) is the most dominant and accounts for 66–95% of the rain. Rain in excess of 50 mm per day is rare in Oman (0.4 and 2.9%), but when it does occur can result in serious consequences such as flash flooding, human catastrophes and land degradation. Rainfall records indicate that Muscat and surrounding areas are susceptible to tropical cyclones and catastrophic rainfall (&gt;100 mm rainfall per day) approximately every 50 years. Copyright © 2008 Royal Meteorological Society

Trends in daily rainfall in the Iberian Peninsula from 1951 to 2002

AbstractAn analysis of 22 sites of daily precipitation records over the period 1951–2002 for the Iberian Peninsula is presented. Annual and seasonal total precipitation (P), number of wet days (N), precipitation intensity (I), the 95th percentile (P95), and percentage of rain falling on days with rainfall above the 95th percentile (%) are investigated. The annual and seasonal trends for these variables and for all 22 rain gauges are analysed, using the Mann–Kendall statistic, and a linear regression model. Moreover, a t‐test is applied to the difference between the means of two subperiods, respectively 1951–1976 and 1977–2002. Principal results indicate a decreasing trend in P, I, and P95 for several northern and southern stations in winter; P, I, and P95 in some southern stations in spring, I and P95 in some southern stations in summer, and I in some northern and southern stations in autumn. The general behaviour is a decrease in the daily intensity of rainfall, while the number of wet days does not reveal pronounced changes. This pattern is valid for both annual and seasonal values of the indices. The decreasing trend found for I in winter and annual series for some localities may be related to the predominance of the positive phase of the North Atlantic Oscillation (NAO), but it is necessary to find other mechanisms for those stations and seasons not linked directly to NAO. Copyright © 2006 Royal Meteorological Society.

REGRESSION-BASED FLOW ADJUSTMENT PROCEDURES FOR TREND ANALYSIS OF WATER QUALITY DATA

Trend analysis is an important component of most total maximum daily load (TMDL) projects in assessing the adequacy of implementation plans on water quality improvement. One problem in trend analysis of water quality data is the confounding effect of varying streamflow conditions at the time of sampling. Commonly used methods of flow adjustment are ordinary least squares (OLS) regression and locally weighted regression and smoothing scatterplots (LOWESS) with a prespecified smoothing factor (f) of 0.5. We compared trend analysis results using these two methods of flow adjustment as well as LOWESS flow adjustment approach with the optimal f value based upon the bias corrected Akaike information criterion (AICc1). The objectives of the study were to evaluate the importance of flow adjustment in trend analysis and to determine the best flow adjustment method among these three procedures. Monthly concentrations from grab and flow-weighted data sets of soluble orthophosphate-phosphorus (PO4-P), total P (TP), and total suspended solid (TSS) were evaluated for three stream sites within the North Bosque River watershed in central Texas. The nonparametric Kendall statistic was employed to test for the presence of trend. As expected, the detection of trend was greatly influenced by flow adjustment. The LOWESS procedure was considered more appropriate than OLS, because LOWESS was able to better define the relationships between flow and constituent concentration. The AICc1 based f value and f = 0.5 in the LOWESS procedure gave similar trend results indicating that the default f value of 0.5 is adequate for reducing variability in constituent concentrations due to flow.

The influence of autocorrelation on the ability to detect trend in hydrological series

AbstractThis study investigated using Monte Carlo simulation the interaction between a linear trend and a lag‐one autoregressive (AR(1)) process when both exist in a time series. Simulation experiments demonstrated that the existence of serial correlation alters the variance of the estimate of the Mann–Kendall (MK) statistic; and the presence of a trend alters the estimate of the magnitude of serial correlation. Furthermore, it was shown that removal of a positive serial correlation component from time series by pre‐whitening resulted in a reduction in the magnitude of the existing trend; and the removal of a trend component from a time series as a first step prior to pre‐whitening eliminates the influence of the trend on the serial correlation and does not seriously affect the estimate of the true AR(1). These results indicate that the commonly used pre‐whitening procedure for eliminating the effect of serial correlation on the MK test leads to potentially inaccurate assessments of the significance of a trend; and certain procedures will be more appropriate for eliminating the impact of serial correlation on the MK test. In essence, it was advocated that a trend first be removed in a series prior to ascertaining the magnitude of serial correlation. This alternative approach and the previously existing approaches were employed to assess the significance of a trend in serially correlated annual mean and annual minimum streamflow data of some pristine river basins in Ontario, Canada. Results indicate that, with the previously existing procedures, researchers and practitioners may have incorrectly identified the possibility of significant trends. Copyright © Environment Canada. Published by John Wiley &amp; Sons, Ltd.

Regional streamflow trend detection with consideration of both temporal and spatial correlation

AbstractIt is known that serial correlation within time series at sites and cross‐correlation among sites in a specific region will influence the ability of statistical tests to assess the field significance of trends over the region. However, serial and/or cross‐correlation has been ignored in field trend‐analyses. This study attempts to develop a methodology that takes into account both serial and cross‐correlation in the assessment of the field significance of trends. The regional average Mann–Kendall (RAMK) statistic is used to represent the regional properties of trends at a regional scale. The null distribution of the RAMK statistic is derived on the basis that the joint probability distribution of m independent normal variables is also normally distributed. The variance of the RAMK statistic is then modified by serial and cross‐correlation. The applicability of the method was demonstrated by applying it to assess the field significance of trends in annual mean, annual maximum, and annual minimum daily streamflow from 1967 to 1996 in ten major homogeneous climate regions of Canada. The results indicate that the method developed provides more accurate assessment of the field significance of trends than that without consideration of serial and cross‐correlation.At the significance level of 0.10, annual mean daily flow increased significantly in the region of Yukon and northern BC mountains whereas it decreased significantly in the Pacific and the Prairie regions. Annual maximum daily flow decreased significantly across southern Canada, except in the Pacific region. Annual minimum daily flow decreased significantly in the Pacific region and in southeastern Canada, with the exception of the region of Great Lakes and St Lawrence river basin, whereas it increased significantly in the region of Yukon and northern BC mountains. Copyright © 2002 Royal Meteorological Society