Texas A&M University

Department of Civil Engineering

College Station, TX, USA.

 

Course : CVEN689

Instructor : Dr. Francisco Olivera

 

Project Title : GIS based Lake/Reservoir evaporation      calculations in the state of Texas, USA.

 

 

Submitted by : Paramjit Chibber

Date : 04/29/02

 

 

 

 

Index

Abstract

Introduction

Literature Review

Methodology

Application, results and discussion`

Conclusions

 

Abstract : The basic purpose of my term project was to calculate with the help of GIS, the annual evaporation (in Acre Ft) from all the reservoirs in the state of Texas ( given the pan evaporation at gage stations) during the years 1958, 1968, 1978, 1988, 1998.The project began with acquiring the the requisite datasets. Every dataset that I needed I got it from the world wide web. The main problem was that of  the pan evaporation data for gage stations that I got from the NWS Hydrologic Research Lab, as it  had some missing values, so those missing values I filled up (for my project only) with the evaporation data (for Quadrangles)  available at TWDB internet site. The methods used and compared were Thiessen polygons and Grid method. In Thiessen polygons, polygons of equal evaporation are generated whereas in Grid method, an evaporation surface was created with the help of interpolation. The results from both the methods showed similar trends  with a strange observation for the year 1968, as in that year the interpolation or grid  method showed some negative values for the evaporation, which certainly is not possible. This observation concluded that conversion to grids or surfaces is not an error free process and for better results with this method we must have sufficient and to an extent uniform placement of point( gage in this case) stations in our study Area. Also it had been observed that Thiessen polygon method gave me  the higher values for evaporation.

Introduction : According to Stephen .A. Thomson in Hydrology for Water Management, "Evaporation is a major water "loss" from reservoirs, especially in arid and semi arid climates like those of western United States. The single largest use of water in the entire state of New Mexico is evaporation from Elephant Butte reservoir on Rio Grande. Lake evaporation exceeds 100 inches per year in the deserts of southeastern California. The pattern is highly variable in the mountain states, but evaporation changes fairly regularly from west to east across the Great plains. In eastern United States evaporation decreases with latitude, with the highest values along the lower Mississippi River and the lowest values following the spine of the Appalachian Mountains into the Northeast".

Evaporation records play a vital role in demand and availability of water from a reservoir. As while making future estimations for the availability of water from a reservoir, the future evaporation besides other things is usually accounted for.

Literature Review TWDB has formulated a GIS program called THEvap, which computes monthly gross evaporation rates.ThEvap produces the monthly evaporation by quadrangle.ThEvap runs on a UNIX workstation and uses updated pan evaporation coefficients to the observed daily evaporation to calculate the monthly reservoir evaporation rates for Texas. In the monthly gross evaporation rates some missing data has been replaced by long term mean values for quadrangle.

In my project rather than taking quadrangles I tried to calculate and compare annual evaporation by two methods viz. Thiessen polygon and Grid Method.

Methodology : 

Step1)The pan evaporation at gage stations dataset obtained from the NWS Hydrologic Research Lab was not complete. So just for my term project only, I used the evaporation data from TWDB internet site to fill up the missing values.TWDB evaporation data is available for quadrangles in the state of Texas. To fill up the missing data I found out which gage station lies in which quadrangle and then filled  the missing/not available data. Before filling the missing data I converted the evaporation data into pan evaporation data, so that all the dataset is in one same format.

Step2)After that I did the evaporation calculations, using pan evaporation coefficients obtained from the TWDB internet site (available for quadrangles) for the years under my study. This way I converted the pan evaporation data into water surface evaporation data at gage stations.

Step3)I made separate tables (as all gage stations did not have the data for all the years under my study)  for each of the years in text format ,followed by adding those tables into the Arc View  and in the end converted them to five different point themes. The need for different point themes arose because not all gage stations have the data for the years under my study , so for each year I have selected only those gage stations which have complete or some data and I did not selected those gage stations which provided me with no information at all. The attribute table for the year 1958 looks as the following :

 

 

Step4) After this I was ready to begin with the evaporation calculations for the reservoirs. For evaporation calculations two methods were adopted:-

   a) Thiessen polygon method.

   b) Grid method.

 

A) Thiessen polygons :-

Step a-1) Before creating the Thiessen polygons I needed the outer boundary of Texas theme, for that I dissolved the Texas counties dataset (obtained from TNRIS). Now when I tried to create Thiessen polygons out of point (gage station) theme, it was found that one of the gage station lies out side the boundary of Texas , so as suggested by our instructor Dr. Francisco Olivera I buffered the boundary of Texas.

Fig-1 : When zoomed to the rectangle, one gage station found outside the boundary of Texas.

Fig 2 : View when zoomed to the rectangle.

 

.

Fig 3) Buffered boundary of Texas, Gage station lies within the boundary.

 

Step a-2)After that Thiessen polygons were created from all point (gage station) themes and were assigned   the values from annual_Yr  field.

Fig 4 :Shows thiessen polygons and its attribute table

 

Step a-3)After that it was observed that the same reservoir can have different evaporation depending upon which part of its falls in which Thiessen polygon.

Figs 5 : When zoomed to the rectangle shows how a reservoir can have different evaporation rates in different parts of its own. 

Step a-4)To account for the above observation, I intersected the reservoirs polygon theme with the Thiessen polygon theme. Doing this way the number of polygons significantly increases.

Step a-5)Now from the CRWR extension , I used Update feature geometry option to update areas of the intersected reservoirs. Then I added a new field Ev(AcFt) i.e. evaporation in Acre field in the attribute of intersected reservoir theme table and  filled this field with :

 Area* Annual_ev_88*247.1057/12000000. 

 

Step a-6)After this I added the intersected parts of the reservoirs using Dissolve option of the extension Geoprocessing Wizard.

Step a-7)Now from the Statistics in the Field menu, I got the sum total of evaporation from all the reservoirs.

This can be seen in the table below.

 

Table 2 : Shows Steps a-4, a-5, a-6, a-7.

 

Step a-7 : Repeated the above steps for all the years to get annual evaporation in those years.

 

B) Grid Method :-

Step b-1)Add buffered Texas outer boundary theme .Add gage stations theme. From  the Analysis menu of the Spatial analyst extension, I used "interpolate grid" for gage station point theme, considering the outer boundary as the buffered outer boundary of Texas. I selected a cell size of 500 mts. and the assigned  value to each cell is from the Annual_88 (say) field. Doing this way I generated a new grid dataset.

 

     

Step b-2)After that I added the reservoir theme .Then from the summarize zones option of the Spatial Analyst extension I got the mean evaporation in each zone or reservoir. Then I added a new field Ev(AcFt) and filled it with :

   Ev(AcFt) = Area*Mean Evaporation*247.1057/12000000.

Step b-3)Now from the Statistics option I got the annual evaporation from all the reservoirs.

 

Step b-4)Repeated the above steps for all the years to get annual evaporation in those years.

 

Application, results and discussion :

1) Study Area : The study area involves the Texas state as a whole and in particular all the reservoirs in it. I concentrated on one single process in the hydrologic cycle i.e. Evaporation. 

2) Datasets Used :

    i)    Pan evaporation data at gage stations from NWS Hydrologic Research Lab.

    ii)   Evaporation data in quadrangles from TWDB.

    iii)  Reservoirs and counties dataset from TNRIS.

    iv)  Pan evaporation coefficient in Quadrangles from TWDB.

3) Results :

1)  Annual Evaporation from Thiessen polygon method.

 

Year :                                   1958           1968           1978          1988          1998

Evaporation (Acre Ft) : 11475752  11462713  14315370 12117721  12029697

 

 

 

 

2)  Annual Evaporation from Grid method.

Year                                1958         1968         1978         1988           1998

Evaporation (AcFt) : 12008612 10766627 12574776 12174579  12070145
 

 

 

4) Inferences :

   i) Reservoir evaporation was maximum during 1978 amongst all the years under study from both the methods.

  ii) Reservoir evaporation was minimum during 1968 amongst the years under study from both the methods.

 iii) The two methods didn't really have close results.

 iv) On comparing the results from both the methods, during the year 1968, a huge difference can be seen in the values of annual evaporation .The reason being, when the annual_68 values were interpolated, I got some negative values for the evaporation in the Northern part of Texas.

 

Conclusions : 

From the above calculations and inferences thereof, it can be said that GIS is a a powerful tool in doing calculations, in which we interpolate a point data to acquire a surface data. Also amongst the two methods discussed above Thiessen polygon method gave me the  better results. This might be a conservative approach, but as seen in the case of year 1968 Grid Method can sometimes give us purely wrong data ( like some negative values for evaporation in this case). In case we want better results with Grid method we must have to an extent uniform placement of point stations and also the cell size should  be small. Thus I feel  that the data, be available at all the gage stations in Texas, the results would have been more precise.

 

 

Class Presentation( Power Point)

 

References :

1) "Hydrology for water management " Stephen A Thomson, Department of Geography,    Millersville University, Pennsylvania.

2) NWS Hydrologic Research Lab.( Pan evaporation data at gage stations. )

3)TWDB. (Evaporation and pan evaporation coefficients data in quadrangles. )

    hyper20.twdb.state.tx.us/Evaporation/evap2.html

 

Special  Thanks to :

1) Dr. Francisco Olivera

    Asstt. Prof., Department of Civil Engineering

    Texas A&M University

     College Station, Texas.

2)NWS Hydrologic Research Lab.

   ( For providing pan evaporation data online )

3)TWDB.

    ( For providing  evaporation and pan evaporation coefficients data online )

4) Mr. Srikanth Koka

    ( For helping me in learning some of the  basics of the Computers)

 

 

Top